Étude de la biodiversité des espaces verts sur le campus de Louvain-la-Neuve: type de gestion et réseaux plantes-pollinisateurs
Jeanne David
04 juin, 2023,19:57
1 Préparation des données
# Chargement des données
# Les .csv sont importés et stockés
Donnees_pres <- read.csv("Data/Donnees_pres_agg.csv", header = T, sep = ";", dec = ",", row.names = 2)
Donnees_pres <- as.data.frame(Donnees_pres)
FinJuil_AllPl <- read.csv("Data/FinjuilletAllplants.csv", header = T, sep = ";", dec = ",")
Interactions <- read.csv2("Data/Interactions.csv")
Site_gestion <- read.csv2("Data/Site_Gestion.csv")
Classes <- read.csv2("Data/Classes_poll.csv")
expe_Tonte <- read.csv("Data/Expe_tonte.csv", header = T, sep = ";", dec = ",")Renommer les colonnes et mise en facteurs des variables
## Données prés
#Site_gestion_date = as.factor(Site_gestion_date),
Donnees_pres <- Donnees_pres %>%
mutate(Site = as.factor(Site),
Gestion_2 = as.factor(Gestion_2),
Parcelle = as.factor(Parcelle),
Gestion_3 = as.factor(Gestion_3),
Gestion_4 = as.factor(Gestion_4),
Mixte_isole = as.factor(Mixte_isole),
Quartier = as.factor(Quartier),
Jours = as.factor(Jours),
Gestion_moment_4 = as.factor(Gestion_moment_4),
Gestion_moment_5 = as.factor(Gestion_moment_5),
Activite = as.factor(Activite),
Periode = as.factor(Periode),
Meteo = as.factor(Meteo))
#colnames(Donnees_pres)
Donnees_pres$Date <- dmy(Donnees_pres$Date)
# Donnees_pres$Heure_debut <- hms(Donnees_pres$Heure_debut)
# Donnees_pres$Heure_fin <- hms(Donnees_pres$Heure_fin)
levels(Donnees_pres$Periode) <- c("Juin", "Mi-juillet", "Fin juillet")
levels(Donnees_pres$Meteo) <- c("Alternances", "Nuageux", "Soleil")
levels(Donnees_pres$Mixte_isole) <- c("Isolé fauche", "Isolé tonte", "Mixte")
levels(Donnees_pres$Gestion_moment_4) <- c("Fauche", "Semé", "Tonte récente", "Tonte tardive")
levels(Donnees_pres$Gestion_moment_5) <- c("Fleuri", "Graminées", "Semé", "Tonte récente", "Tonte tardive")
## Inventaire
#Site_gestion_date,
Inventaire <- Donnees_pres %>%
select(Site,
Nombre_quadrats,
Gestion_2, Parcelle, Gestion_4,
Mixte_isole,
Area_gis_m_sq,
#Green100, Building100, Impervious100,Natural100
Quartier,
Jours,
Gestion_moment_4, Gestion_moment_5,
Activite,
Periode,
Date,
Heure_debut, Heure_fin,
Temperature, Meteo,
Achillea_millefolium:Vicia_tetrasperma_subsp._tetrasperma,
Aglais_io:Volucella_zonaria)
Inventaire$Gestion_4 <- fct_relevel(Inventaire$Gestion_4, c("Graminees", "Fleuri", "Seme", "Tonte"))
Inventaire$Gestion_moment_5 <- fct_relevel(Inventaire$Gestion_moment_5, c("Graminées", "Fleuri", "Semé", "Tonte récente", "Tonte tardive"))
Inventaire$Mixte_isole <- fct_relevel(Inventaire$Mixte_isole, c("Isolé fauche", "Mixte", "Isolé tonte"))
Inventaire$Activite <- fct_relevel(Inventaire$Activite, c("Nulle", "Moyenne", "Forte"))
Inventaire$Meteo <- fct_relevel(Inventaire$Meteo, c("Nuageux", "Alternances", "Soleil"))
## Fin juillet all plants
FinJuil_AllPl <- FinJuil_AllPl %>%
mutate(Site = as.factor(Site),
Gestion_2 = as.factor(Gestion_2),
Parcelle = as.factor(Parcelle),
Gestion_4 = as.factor(Gestion_4),
Mixte_isole = as.factor(Mixte_isole),
Quartier = as.factor(Quartier),
Jours = as.factor(Jours),
Gestion_moment_4 = as.factor(Gestion_moment_4),
Gestion_moment_5 = as.factor(Gestion_moment_5),
Activite = as.factor(Activite),
Periode = as.factor(Periode),
Meteo = as.factor(Meteo))
FinJuil_AllPl$Date <- dmy(FinJuil_AllPl$Date)
FinJuil_AllPl$Heure_debut <- hms(FinJuil_AllPl$Heure_debut)
FinJuil_AllPl$Heure_fin <- hms(FinJuil_AllPl$Heure_fin)
levels(FinJuil_AllPl$Periode) <- c("Fin juillet")
levels(FinJuil_AllPl$Meteo) <- c("Alternances", "Nuageux", "Soleil")
levels(FinJuil_AllPl$Mixte_isole) <- c("Isolé fauche", "Isolé tonte", "Mixte")
levels(FinJuil_AllPl$Gestion_moment_4) <- c("Fauche", "Semé", "Tonte récente", "Tonte tardive")
levels(FinJuil_AllPl$Gestion_moment_5) <- c("Fleuri", "Graminées", "Semé", "Tonte récente", "Tonte tardive")
AllFinJuillet <- FinJuil_AllPl %>%
select(Site,
nombre_quadrats,
Gestion_2, Parcelle, Gestion_4,
Mixte_isole,
Area,
#Green100, Building100, Impervious100,Natural100
Quartier,
Jours,
Gestion_moment_4, Gestion_moment_5,
Activite,
Periode,
Date,
Heure_debut, Heure_fin,
Temperature, Meteo,
Achillea_millefolium:Vicia_tetrasperma_subsp._tetrasperma,
Aglais_io:Volucella_zonaria)
AllFinJuillet$Gestion_4 <- fct_relevel(AllFinJuillet$Gestion_4, c("Graminees", "Fleuri", "Seme", "Tonte"))
AllFinJuillet$Gestion_moment_5 <- fct_relevel(AllFinJuillet$Gestion_moment_5, c("Graminées", "Fleuri", "Semé", "Tonte récente", "Tonte tardive"))
## Interactions
Interactions <- Interactions %>%
mutate(Site_gestion_date_Quadrat = as.factor(Site_gestion_date_Quadrat),
Sp_Plantes = as.factor(Sp_Plantes),
Sp_Pollinisateurs = as.factor(Sp_Pollinisateurs))
## Site & Classes
colnames(Site_gestion)[1:2] <- c("Site_gestion_date_Quadrat", "Site_gestion_date")
Site_gestion <- Site_gestion %>%
mutate(Site_gestion_date_Quadrat = as.factor(Site_gestion_date_Quadrat),
Site_gestion_date = as.factor(Site_gestion_date))
colnames(Classes)[1] <- c("Sp_Pollinisateurs")
Classes <- Classes %>%
mutate(Sp_Pollinisateurs = as.factor(Sp_Pollinisateurs) ,
Classe_Poll = as.factor(Classe_Poll))2 Illustration et survol des données
2.1 Inventaires plantes et pollinisateurs
Définir une colonne Diversité spécifique (nombre d’espèces, richesse spécifique S) et Abondance des plantes et des pollinisateurs
Inventaire$S_Plant <- specnumber(Inventaire[19:67])
Inventaire$Ab_Plant <- rowSums(Inventaire[19:67])
Inventaire$S_Poll <- specnumber(Inventaire[68:166])
Inventaire$Ab_Poll <- rowSums(Inventaire[68:166])Jeu de données simplifié et corrélations
Inv <- Inventaire[,c(1:18,167:170)]
corrplot(cor(Inv[,c(7,17,19:22)]), order = "hclust", type = "upper", tl.col = "black")
2.1.1 Statistiques descriptives
2.1.1.1 Graphiques
2.1.1.1.1 Types de gestion
Inv %>% ggplot(aes (x = Gestion_2, y = S_Plant, color = Gestion_2)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
labs(#title = "Richesse spécifique en plantes en fonction des types de gestion",
x = "", y = "Richesse spécifique en plantes") +
theme(legend.position = "none") +
Inv %>% ggplot(aes (x = Gestion_2, y = Ab_Plant, color = Gestion_2)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
labs(#title = "Abondance en plantes en fonction des types de gestion",
x = "", y = "Abondance en plantes") +
theme(legend.position = "none") +
Inv %>% ggplot(aes (x = Gestion_2, y = S_Poll, color = Gestion_2)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
labs(#title = "Richesse spécifique en pollinisateurs en fonction des types de gestion",
x = "Type de gestion", y = "Richesse spécifique en pollinisateurs") +
theme(legend.position = "none") +
Inv %>% ggplot(aes (x = Gestion_2, y = Ab_Poll, color = Gestion_2)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
labs(#title = "Abondance en pollinisateurs en fonction des types de gestion",
x = "Type de gestion", y = "Abondance en pollinisateurs") +
theme(legend.position = "none")
Inv %>% ggplot(aes (x = Gestion_4, y = S_Plant)) +
geom_boxplot(alpha = 0.70) +
labs(#title = "Richesse spécifique en plantes en fonction des types de gestion",
x = "", y = "Richesse spécifique en plantes") +
Inv %>% ggplot(aes (x = Gestion_4, y = Ab_Plant)) +
geom_boxplot(alpha = 0.70) +
labs(#title = "Abondance de plantes en fonction des types de gestion",
x = "", y = "Abondance en plantes") +
Inv %>% ggplot(aes (x = Gestion_4, y = S_Poll)) +
geom_boxplot(alpha = 0.70) +
labs(#title = "Richesse spécifique en pollinisateurs en fonction des types de gestion",
x = "Type de gestion", y = "Richesse spécifique en pollinisateurs") +
Inv %>% ggplot(aes (x = Gestion_4, y = Ab_Poll)) +
geom_boxplot(alpha = 0.70) +
labs(#title = "Abondance de pollinisateurs en fonction des types de gestion",
x = "Type de gestion", y = "Abondance en pollinisateurs")
Inv %>% ggplot(aes (x = Gestion_moment_4, y = S_Plant)) +
geom_boxplot(alpha = 0.70) +
labs(#title = "Richesse spécifique en plantes en fonction des types de gestion",
x = "", y = "Richesse spécifique en plantes") +
theme(legend.position="none") +
Inv %>% ggplot(aes (x = Gestion_moment_4, y = Ab_Plant)) +
geom_boxplot(alpha = 0.70) +
labs(#title = "Abondance de plantes en fonction des types de gestion",
x = "", y = "Abondance en plantes") +
theme(legend.position="none") +
Inv %>% ggplot(aes (x = Gestion_moment_4, y = S_Poll)) +
geom_boxplot(alpha = 0.70) +
labs(#title = "Richesse spécifique en pollinisateurs en fonction des types de gestion",
x = "Type de gestion", y = "Richesse spécifique en pollinisateurs", color = "Type de gestion") +
theme(legend.position="bottom") +
Inv %>% ggplot(aes (x = Gestion_moment_4, y = Ab_Poll)) +
geom_boxplot(alpha = 0.70) +
labs(#title = "Abondance de pollinisateurs en fonction des types de gestion",
x = "Type de gestion", y = "Abondance en pollinisateurs", color = "Type de gestion") +
theme(legend.position="none")
Inv %>% ggplot(aes (x = Gestion_moment_5, y = S_Plant)) +
geom_boxplot(aes (color = Gestion_moment_5), alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",#6dcf20
"Tonte tardive" = "#2b790c")) + #12661f
labs(#title = "Richesse spécifique en plantes en fonction des types de gestion",
x = "", y = "Richesse spécifique en plantes") +
theme(legend.position = "none") +
Inv %>% ggplot(aes (x = Gestion_moment_5, y = Ab_Plant)) +
geom_boxplot(aes (color = Gestion_moment_5), alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(#title = "Abondance de plantes en fonction des types de gestion",
x = "", y = "Abondance en plantes") +
theme(legend.position = "none") +
Inv %>% ggplot(aes (x = Gestion_moment_5, y = S_Poll, color = Gestion_moment_5)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(#title = "Richesse spécifique en pollinisateurs en fonction des types de gestion",
x = "Type de gestion", y = "Richesse spécifique en pollinisateurs") +
theme(legend.position = "none") +
Inv %>% ggplot(aes (x = Gestion_moment_5, y = Ab_Poll, color = Gestion_moment_5)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(#title = "Abondance de pollinisateurs en fonction des types de gestion",
x = "Type de gestion", y = "Abondance en pollinisateurs") +
theme(legend.position = "none")
2.1.1.1.2 Selon la période
Inv %>% ggplot(aes (x = Periode, y = S_Plant)) +
geom_boxplot(aes (color = Periode), alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(#title = "Richesse spécifique en plantes en fonction de la période",
x = "", y = "Richesse spécifiqu en plantes") +
theme(legend.position = "none") +
Inv %>% ggplot(aes (x = Periode, y = Ab_Plant)) +
geom_boxplot(aes (color = Periode), alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(#title = "Abondance de plantes en fonction de la période",
x = "", y = "Abondance en plantes") +
theme(legend.position = "none") +
Inv %>% ggplot(aes (x = Periode, y = S_Poll)) +
geom_boxplot(aes (color = Periode), alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(#title = "Richesse spécifique en pollinisateurs en fonction de la période",
x = "Période", y = "Richesse spécifique en pollinisateurs") +
theme(legend.position = "none") +
Inv %>% ggplot(aes (x = Periode, y = Ab_Poll)) +
geom_boxplot(aes (color = Periode), alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(#title = "Abondance de pollinisateurs en fonction de la période",
x = "Période", y = "Abondance en pollinisateurs") +
theme(legend.position = "none") 
Inv %>% ggplot(aes (x = Gestion_moment_5, y = S_Plant)) +
geom_boxplot(aes (color = Periode), alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(#title = "Richesse spécifique en plantes en fonction des types de gestion",
x = "", y = "Richesse spécifique en plantes") +
theme(legend.position="none") +
Inv %>% ggplot(aes (x = Gestion_moment_5, y = Ab_Plant, color = Periode)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(#title = "Abondance de plantes en fonction des types de gestion",
x = "", y = "Abondance en plantes") +
theme(legend.position="none") +
Inv %>% ggplot(aes (x = Gestion_moment_5, y = S_Poll, color = Periode)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(#title = "Richesse spécifique en pollinisateurs en fonction des types de gestion",
x = "Type de gestion", y = "Richesse spécifique en pollinisateurs") +
theme(legend.position="none") +
Inv %>% ggplot(aes (x = Gestion_moment_5, y = Ab_Poll, color = Periode)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(#title = "Abondance de pollinisateurs en fonction des types de gestion",
x = "Type de gestion", y = "Abondance en pollinisateurs", color = "Période") +
theme(legend.position="bottom") 
Inv %>% ggplot(aes (x = Periode, y = S_Plant, color = Gestion_moment_5)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(#title = "Richesse spécifique en plantes en fonction des types de gestion",
x = "", y = "Richesse spécifique en plantes") +
theme(legend.position="none") +
Inv %>% ggplot(aes (x = Periode, y = Ab_Plant, color = Gestion_moment_5)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(#title = "Abondance de plantes en fonction des types de gestion",
x = "", y = "Abondance en plantes") +
theme(legend.position="none") +
Inv %>% ggplot(aes (x = Periode, y = S_Poll, color = Gestion_moment_5)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(#title = "Richesse spécifique en pollinisateurs en fonction des types de gestion",
x = "Période", y = "Richesse spécifique en pollinisateurs") +
theme(legend.position="none") +
Inv %>% ggplot(aes (x = Periode, y = Ab_Poll, color = Gestion_moment_5)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(#title = "Abondance de pollinisateurs en fonction des types de gestion",
x = "Période", y = "Abondance en pollinisateurs", color= "Type de gestion") +
theme(legend.position="bottom")
#esquisse::esquisser()
Inv_tib <- as_tibble(Inv)
resum_invG2 <- Inv_tib %>%
group_by(Periode, Gestion_2) %>%
summarize(mean = mean(S_Plant),
var=var(S_Plant),
N=n()) %>%
mutate(CI_S=mean+qt(0.975, N-1)*sqrt(var/N),
CI_I=mean-qt(0.975, N-1)*sqrt(var/N))
ggplot(resum_invG2, aes(x= Periode, y=mean, colour= Gestion_2, group = Gestion_2)) +
geom_errorbar(aes(ymin=CI_I, ymax=CI_S),width=.15,size = 1) +
geom_point(size= 3) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
geom_line(size = 1)+
labs(x = "Période",
y = "Richesse spécifique en plantes",
color = "Type de gestion") 
resum_inv_SPl <- Inv_tib %>%
group_by(Periode, Gestion_moment_5) %>%
summarize(mean = mean(S_Plant),
var=var(S_Plant),
N=n()) %>%
mutate(CI_S=mean+qt(0.975, N-1)*sqrt(var/N),
CI_I=mean-qt(0.975, N-1)*sqrt(var/N))
resum_inv_APl <- Inv_tib %>%
group_by(Periode, Gestion_moment_5) %>%
summarize(mean = mean(Ab_Plant),
var=var(Ab_Plant),
N=n()) %>%
mutate(CI_S=mean+qt(0.975, N-1)*sqrt(var/N),
CI_I=mean-qt(0.975, N-1)*sqrt(var/N))
resum_inv_SPo <- Inv_tib %>%
group_by(Periode, Gestion_moment_5) %>%
summarize(mean = mean(S_Poll),
var=var(S_Poll),
N=n()) %>%
mutate(CI_S=mean+qt(0.975, N-1)*sqrt(var/N),
CI_I=mean-qt(0.975, N-1)*sqrt(var/N))
resum_inv_APo <- Inv_tib %>%
group_by(Periode, Gestion_moment_5) %>%
summarize(mean = mean(Ab_Poll),
var=var(Ab_Poll),
N=n()) %>%
mutate(CI_S=mean+qt(0.975, N-1)*sqrt(var/N),
CI_I=mean-qt(0.975, N-1)*sqrt(var/N))
ggplot(resum_inv_SPl, aes(x= Periode, y=mean, colour= Gestion_moment_5, group = Gestion_moment_5)) +
geom_errorbar(aes(ymin=CI_I, ymax=CI_S),width=.15,size = 1) +
geom_point(size= 3) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
geom_line(size = 1)+
labs(x = "Période",
y = "Richesse spécifique en plantes",
color = "Type de gestion") +
theme(legend.position="none") +
ggplot(resum_inv_APl, aes(x= Periode, y=mean, colour= Gestion_moment_5, group = Gestion_moment_5)) +
geom_errorbar(aes(ymin=CI_I, ymax=CI_S),width=.15,size = 1) +
geom_point(size= 3) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
geom_line(size = 1)+
labs(x = "Période",
y = "Abondance en plantes",
color = "Type de gestion") +
theme(legend.position="none") +
ggplot(resum_inv_SPo, aes(x= Periode, y=mean, colour= Gestion_moment_5, group = Gestion_moment_5)) +
geom_errorbar(aes(ymin=CI_I, ymax=CI_S),width=.15,size = 1) +
geom_point(size= 3) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
geom_line(size = 1)+
labs(x = "Période",
y = "Richesse spécifique en pollinisateurs",
color = "Type de gestion") +
theme(legend.position="none") +
ggplot(resum_inv_APo, aes(x= Periode, y=mean, colour= Gestion_moment_5, group = Gestion_moment_5)) +
geom_errorbar(aes(ymin=CI_I, ymax=CI_S),width=.15,size = 1) +
geom_point(size= 3) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
geom_line(size = 1)+
labs(x = "Période",
y = "Abondance en pollinisateurs",
color = "Type de gestion") +
theme(legend.position="bottom")
2.1.1.1.3 Influences multi-factorielles
2.1.1.1.3.1 Météo
ggplot(Inv) +
aes(x = Meteo, y = Ab_Poll, color = Meteo) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Nuageux" = "#7f908c",
"Alternances" = "#79ccbd",
"Soleil" = "#fbcb09")) +
labs(x = "Météo",
y = "Abondance en pollinisateurs") +
theme(legend.position="none")
ggplot(Inv) +
aes(x = Gestion_moment_5, y = Ab_Poll, color = Meteo) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Nuageux" = "#7f908c",
"Alternances" = "#79ccbd",
"Soleil" = "#fbcb09")) +
labs(x = "Type de gestion",
y = "Abondance en pollinisateurs",
color = "Météo") +
theme(legend.position="bottom")
2.1.1.1.3.2 Températures
ggplot(Inv) +
aes(x = Temperature, y = Ab_Poll) +
geom_point()+
geom_smooth(se = T) +
labs(x = "Température",
y = "Abondance en pollinisateurs")
ggplot(Inv) +
aes(x = Temperature, y = Ab_Poll, color = Gestion_moment_5) +
geom_point()+
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
geom_smooth(se = F) +
labs(x = "Température",
y = "Abondance en pollinisateurs",
color = "Type de gestion") +
theme(legend.position="bottom")
2.1.1.1.3.3 Activité anthropiques
ggplot(Inv) +
aes(x = Activite, y = S_Poll) +
geom_boxplot(alpha = 0.70)+
labs(x = "Activité",
y = "Richesse spécifique en pollinisateurs") +
ggplot(Inv) +
aes(x = Activite, y = Ab_Poll) +
geom_boxplot(alpha = 0.70)+
labs(x = "Activité",
y = "Abondance en pollinisateurs")
ggplot(Inv) +
aes(x = Activite, y = Ab_Poll, color = Gestion_moment_5) +
geom_boxplot(alpha = 0.70)+
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Activité",
y = "Abondance en pollinisateurs",
color = "Type de gestion") +
theme(legend.position="bottom")
ggplot(Inv) +
aes(x = Activite, y = S_Plant) +
geom_boxplot(alpha = 0.70)+
labs(x = "Activité",
y = "Richesse spécifique en plantes")+
ggplot(Inv) +
aes(x = Activite, y = Ab_Plant) +
geom_boxplot(alpha = 0.70)+
labs(x = "Activité",
y = "Abondance en plantes")
ggplot(Inv) +
aes(x = Activite, y = Ab_Plant, color = Gestion_moment_5) +
geom_boxplot(alpha = 0.70)+
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Activité",
y = "Abondance en plantes",
color = "Type de gestion") +
theme(legend.position="bottom")
2.1.1.1.3.4 Heure de la journée
ggplot(Inv) +
aes(x = Heure_debut, y = Ab_Poll, color = Gestion_2) +
geom_point() +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
scale_x_discrete(breaks = c("09:00:00","10:00:00","11:00:00","12:00:00","13:00:00","14:00:00","15:00:00","16:00:00", "17:00:00"),
#limits = c("09:00:00","10:00:00","11:00:00","12:00:00","13:00:00","14:00:00","15:00:00","16:00:00", "17:00:00")
) +
labs(x = "Heure",
y = "Abondance en pollinisateurs") 
ggplot(Inv) +
aes(x = Heure_debut, y = Ab_Poll, color = Gestion_moment_5) +
geom_point()+
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
scale_x_discrete(breaks = c("09:00:00","10:00:00","11:00:00","12:00:00","13:00:00","14:00:00","15:00:00","16:00:00", "17:00:00"),
#limits = c("09:00:00","10:00:00","11:00:00","12:00:00","13:00:00","14:00:00","15:00:00","16:00:00", "17:00:00")
) +
labs(x = "Heure",
y = "Abondance en pollinisateurs",
color = "Type de gestion") +
theme(legend.position="bottom")
ggplot(Inv) +
aes(x = Heure_fin, y = Ab_Poll, color = Gestion_2) +
geom_point() +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
scale_x_discrete(breaks = c("09:30:00","10:30:00","11:30:00","12:30:00","13:30:00","14:30:00","15:35:00","16:30:00", "17:30:00"),
#limits = c("09:30:00","10:30:00","11:30:00","12:30:00","13:30:00","14:30:00","15:30:00","16:30:00", "17:30:00")
) +
labs(x = "Heure",
y = "Abondance en pollinisateurs") 
ggplot(Inv) +
aes(x = Heure_fin, y = Ab_Poll, color = Gestion_moment_5) +
geom_point()+
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
scale_x_discrete(breaks = c("09:30:00","10:30:00","11:30:00","12:30:00","13:30:00","14:30:00","15:35:00","16:30:00", "17:30:00"),
#limits = c("09:30:00","10:30:00","11:30:00","12:30:00","13:30:00","14:30:00","15:30:00","16:30:00", "17:30:00")
) +
labs(x = "Heure",
y = "Abondance en pollinisateurs",
color = "Type de gestion") +
theme(legend.position="bottom")
2.1.1.2 Statistiques par groupes
Création d’une fonction pour faire des statistiques descriptives par groupe
statdesbygroup=function(Y, X)
{# 1. On récupére les niveaux de X et le nombre de niveaux
X=as.factor(X)
levX=levels(X)
nlevX=length(levX)
# 2. On crée la matrice pour les statistique descriptive
mesStats<-matrix(nrow=nlevX,ncol=7)
colnames(mesStats)=c("N","Moyenne","Médiane","Ecart-type","Variance","Min","Max" )
rownames(mesStats)=levels(Inv$Gestion_moment_5)
# 3. On calcule les statistiques descriptives
for(i in 1:nlevX)
{indlevel=which(X==levX[i])
mesStats[i,]<-c(length(Y[indlevel]),mean(Y[indlevel]),median(Y[indlevel]),sd(Y[indlevel]),var(Y[indlevel]),min(Y[indlevel]), max(Y[indlevel]))}
# 4. On retourne les résultats
return(mesStats)
}
SBG_SPlant <- statdesbygroup(
Y = c(Inv$S_Plant),
X = c(Inv$Gestion_moment_5))
SBG_SPlant ## N Moyenne Médiane Ecart-type Variance Min Max
## Graminées 11 2.090909 2 2.0225996 4.0909091 0 5
## Fleuri 30 5.566667 5 2.9674479 8.8057471 1 13
## Semé 18 9.055556 9 3.6213781 13.1143791 3 17
## Tonte récente 26 1.923077 2 0.9347974 0.8738462 0 4
## Tonte tardive 49 3.510204 3 1.3711681 1.8801020 1 6# %>%
# kbl(caption = "Richesse spécifique de la flore") %>%
# kable_classic(full_width = F, html_font = "Cambria") # %>% kable_styling() %>%
# save_kable(file = "Output/Tableau/table_SPlant.pdf")
SBG_AbPlant <- statdesbygroup(
Y = c(Inv$Ab_Plant),
X = c(Inv$Gestion_moment_5))
SBG_AbPlant ## N Moyenne Médiane Ecart-type Variance Min Max
## Graminées 11 18.72727 15.0 18.81006 353.8182 0 52
## Fleuri 30 101.96667 49.0 114.09297 13017.2057 3 439
## Semé 18 131.11111 110.5 81.39438 6625.0458 11 299
## Tonte récente 26 30.34615 21.5 27.44586 753.2754 0 90
## Tonte tardive 49 90.63265 65.0 84.70874 7175.5706 7 451# %>%
# kbl(caption = "Abondance de la flore") %>%
# kable_classic(full_width = F, html_font = "Cambria") # %>% kable_styling() %>%
# save_kable(file = "Output/Tableau/table_AbPlant.pdf")
SBG_SPoll <- statdesbygroup(
Y = c(Inv$S_Poll),
X = c(Inv$Gestion_moment_5))
SBG_SPoll## N Moyenne Médiane Ecart-type Variance Min Max
## Graminées 11 4.545455 4.0 4.321195 18.672727 0 12
## Fleuri 30 8.733333 8.5 4.143406 17.167816 0 17
## Semé 18 12.055556 12.0 4.165294 17.349673 4 21
## Tonte récente 26 1.538462 1.0 1.272188 1.618462 0 5
## Tonte tardive 49 3.244898 3.0 1.984635 3.938776 0 9# %>%
# kbl(caption = "Richesse spécifique de la faune pollinisatrice") %>%
# kable_classic(full_width = F, html_font = "Cambria") # %>% kable_styling() %>%
# save_kable(file = "Output/Tableau/table_SPoll.pdf")
SBG_AbPoll <- statdesbygroup(
Y = c(Inv$Ab_Poll),
X = c(Inv$Gestion_moment_5))
SBG_AbPoll ## N Moyenne Médiane Ecart-type Variance Min Max
## Graminées 11 9.454545 4.0 12.532866 157.072727 0 42
## Fleuri 30 18.400000 17.5 11.278909 127.213793 0 43
## Semé 18 34.333333 35.0 14.868047 221.058824 4 61
## Tonte récente 26 2.038462 1.0 1.865063 3.478462 0 7
## Tonte tardive 49 7.469388 6.0 6.010761 36.129252 0 26# %>%
# kbl(caption = "Abondance de la faune pollinisatrice") %>%
# kable_classic(full_width = F, html_font = "Cambria") # %>% kable_styling() %>%
# save_kable(file = "Output/Tableau/table_AbPoll.pdf")statdesbygroup=function(Y, X)
{# 1. On récupére les niveaux de X et le nombre de niveaux
X=as.factor(X)
levX=levels(X)
nlevX=length(levX)
# 2. On crée la matrice pour les statistique descriptive
mesStats<-matrix(nrow=nlevX,ncol=7)
colnames(mesStats)=c("N","Moyenne","Médiane","Ecart-type","Variance","Min","Max" )
rownames(mesStats)=levels(Inv$Periode)
# 3. On calcule les statistiques descriptives
for(i in 1:nlevX)
{indlevel=which(X==levX[i])
mesStats[i,]<-c(length(Y[indlevel]),mean(Y[indlevel]),median(Y[indlevel]),sd(Y[indlevel]),var(Y[indlevel]),min(Y[indlevel]), max(Y[indlevel]))}
# 4. On retourne les résultats
return(mesStats)
}
SBG_SPlant_Periode <- statdesbygroup(
Y = c(Inv$S_Plant),
X = c(Inv$Periode))
SBG_SPlant_Periode ## N Moyenne Médiane Ecart-type Variance Min Max
## Juin 46 4.782609 4 3.450688 11.907246 0 17
## Mi-juillet 46 4.108696 3 3.121455 9.743478 0 13
## Fin juillet 42 3.952381 3 2.819379 7.948897 0 13# %>%
# kbl(caption = "Richesse spécifique de la flore") %>%
# kable_classic(full_width = F, html_font = "Cambria") # %>% kable_styling() %>%
# save_kable(file = "Output/Tableau/table_SPlant_Periode.pdf")
SBG_AbPlant_Periode <- statdesbygroup(
Y = c(Inv$Ab_Plant),
X = c(Inv$Periode))
SBG_AbPlant_Periode ## N Moyenne Médiane Ecart-type Variance Min Max
## Juin 46 119.36957 86.0 111.70484 12477.971 0 451
## Mi-juillet 46 67.41304 37.5 73.57553 5413.359 0 299
## Fin juillet 42 53.88095 37.0 53.42248 2853.961 0 239# %>%
# kbl(caption = "Abondance de la flore") %>%
# kable_classic(full_width = F, html_font = "Cambria") # %>% kable_styling() %>%
# save_kable(file = "Output/Tableau/table_AbPlant_Periode.pdf")
SBG_SPoll_Periode <- statdesbygroup(
Y = c(Inv$S_Poll),
X = c(Inv$Periode))
SBG_SPoll_Periode ## N Moyenne Médiane Ecart-type Variance Min Max
## Juin 46 6.217391 5.5 4.939244 24.39614 0 21
## Mi-juillet 46 5.413043 4.0 5.144905 26.47005 0 19
## Fin juillet 42 4.595238 3.0 3.876505 15.02729 0 16# %>%
# kbl(caption = "Richesse spécifique de la faune pollinisatrice") %>%
# kable_classic(full_width = F, html_font = "Cambria")# %>% kable_styling() %>%
# save_kable(file = "Output/Tableau/table_SPoll_Periode.pdf")
SBG_AbPoll_Periode <- statdesbygroup(
Y = c(Inv$Ab_Poll),
X = c(Inv$Periode))
SBG_AbPoll_Periode ## N Moyenne Médiane Ecart-type Variance Min Max
## Juin 46 13.391304 10.5 11.73689 137.7546 0 41
## Mi-juillet 46 14.478261 8.0 17.16552 294.6551 0 61
## Fin juillet 42 9.785714 4.0 10.66916 113.8310 0 42# %>%
# kbl(caption = "Abondance de la faune pollinisatrice") %>%
# kable_classic(full_width = F, html_font = "Cambria") # %>% kable_styling() %>%
# save_kable(file = "Output/Tableau/table_AbPoll_Periode.pdf")2.1.2 Courbes d’accumulation
#Préparations données
especes <- Inventaire[,c(19:166)]
Esp_Plant <- Inventaire[,c(19:67)]
Esp_Poll <- Inventaire[,c(68:166)]
Inv_red <- Inventaire[,c(3:5,10:11,13:14,167:170)]2.1.2.1 Plantes
# accumcomp(Esp_Plant, y = Inv_red, factor = "Gestion_2", col = c("#aa1e0f","#12661f"), rainbow = F, xlim = c(1,90), plotit = TRUE, labelit = F, legend = F, xlab = "Nombre d'échantillonnages", ylab = "Nombre d'espèces")[1]
# legend("topright", legend = c("Fauche", "Tonte"), col = c("#aa1e0f", "#12661f"), lty = 1, xpd = TRUE, inset = c(0,-0.2), horiz = F)
Accum.Plantes_G2 <- accumcomp(Esp_Plant, y=Inv_red, factor='Gestion_2',
method='exact', conditioned=FALSE, plotit=FALSE)
accum.long_Plantes_G2 <- accumcomp.long(Accum.Plantes_G2, ci=NA, label.freq=1)
plG2 <- ggplot(data=accum.long_Plantes_G2, aes(x = Sites, y = Richness, ymax = UPR, ymin = LWR)) +
scale_x_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
scale_y_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
geom_line(aes(colour=Grouping), size=1.5) +
geom_point(data=subset(accum.long_Plantes_G2, labelit==TRUE),
aes(colour=Grouping, shape=Grouping), size=2.5) +
scale_shape_manual(values = c("Fauche" = 16,
"Tonte" = 5)) +
geom_ribbon(aes(colour=Grouping, fill=after_scale(alpha(colour, alpha=0.15))),
show.legend=FALSE) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
labs(x = "Nombre d'échantillonages", y = "Nombre d'espèces de plantes", colour = "Type de gestion", shape = "Type de gestion") +
theme(legend.position = "bottom")
# accumcomp(Esp_Plant, y = Inv_red, factor = "Gestion_moment_5", col = c("darkgreen", "green"), xlim = c(1,70), plotit = T, labelit = F, rainbow = F, legend = F, xlab = "Nombre d'échantillonages", ylab = "Nombre d'espèces")[1]
# legend("topright", legend = c("Fleuri", "Graminées", "Semé","Tonte récente", "Tonte tardive"), col = rainbow(5), lty = 1, xpd = TRUE, inset = c(0,-0.2), horiz = F)
Accum.Plantes_G5 <- accumcomp(Esp_Plant, y=Inv_red, factor='Gestion_moment_5',
method='exact', conditioned=FALSE, plotit=FALSE)
accum.long_Plantes_G5 <- accumcomp.long(Accum.Plantes_G5, ci=NA, label.freq=1)
accum.long_Plantes_G5$Grouping <- fct_relevel(accum.long_Plantes_G5$Grouping,c("Graminées", "Fleuri", "Semé", "Tonte récente", "Tonte tardive"))
plG5 <- ggplot(data=accum.long_Plantes_G5, aes(x = Sites, y = Richness, ymax = UPR, ymin = LWR)) +
scale_x_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
scale_y_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
geom_line(aes(colour=Grouping), size=1.5) +
geom_point(data=subset(accum.long_Plantes_G5, labelit==T),
aes(colour=Grouping, shape=Grouping), size=2.5) +
scale_shape_manual(values = c("Graminées" = 16,
"Fleuri" = 17,
"Semé" = 15,
"Tonte récente" = 5,
"Tonte tardive" = 6)) +
geom_ribbon(aes(colour=Grouping, fill=after_scale(alpha(colour, alpha=0.15))),
show.legend=FALSE) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Nombre d'échantillonages", y = "Nombre d'espèces de plantes", colour = "Type de gestion", shape = "Type de gestion") +
theme(legend.position = "bottom")
plG2 + plG5
2.1.2.1.1 Périodes
# Juin
Inv_Juin <- Inv_red %>%
filter(Periode == "Juin")
Esp_Plant_Juin <- Esp_Plant %>%
rownames_to_column(var = "temp") %>%
filter(grepl("Juin$", temp)) %>%
column_to_rownames(var = "temp")
Accum.Plantes_Juin <- accumcomp(Esp_Plant_Juin, y=Inv_Juin, factor='Gestion_moment_5',
method='exact', conditioned=FALSE, plotit=FALSE)
accum.long_Plantes_Juin <- accumcomp.long(Accum.Plantes_Juin, ci=NA, label.freq=1)
accum.long_Plantes_Juin$Grouping <- fct_relevel(accum.long_Plantes_Juin$Grouping,c("Graminées", "Fleuri", "Semé", "Tonte récente", "Tonte tardive"))
pl_Juin <- ggplot(data=accum.long_Plantes_Juin, aes(x = Sites, y = Richness, ymax = UPR, ymin = LWR)) +
scale_x_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
scale_y_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
geom_line(aes(colour=Grouping), size=1.5) +
geom_point(data=subset(accum.long_Plantes_Juin, labelit==T),
aes(colour=Grouping, shape=Grouping), size=2.5) +
scale_shape_manual(values = c("Graminées" = 16,
"Fleuri" = 17,
"Semé" = 15,
"Tonte récente" = 5,
"Tonte tardive" = 6)) +
geom_ribbon(aes(colour=Grouping, fill=after_scale(alpha(colour, alpha=0.15))),
show.legend=FALSE) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "", y = "Nombre d'espèces de plantes", colour = "Type de gestion", shape = "Type de gestion") +
theme(legend.position = "none")
# mi Juillet
Inv_miJuillet <- Inv_red %>%
filter(Periode == "Mi-juillet")
Esp_Plant_miJuillet <- Esp_Plant %>%
rownames_to_column(var = "temp") %>%
filter(grepl("miJuillet$", temp)) %>%
column_to_rownames(var = "temp")
Accum.Plantes_miJuillet <- accumcomp(Esp_Plant_miJuillet, y=Inv_miJuillet, factor='Gestion_moment_5',
method='exact', conditioned=FALSE, plotit=FALSE)
accum.long_Plantes_miJuillet <- accumcomp.long(Accum.Plantes_miJuillet, ci=NA, label.freq=1)
accum.long_Plantes_miJuillet$Grouping <- fct_relevel(accum.long_Plantes_miJuillet$Grouping,c("Graminées", "Fleuri", "Semé", "Tonte récente", "Tonte tardive"))
pl_miJuillet <- ggplot(data=accum.long_Plantes_miJuillet, aes(x = Sites, y = Richness, ymax = UPR, ymin = LWR)) +
scale_x_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
scale_y_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
geom_line(aes(colour=Grouping), size=1.5) +
geom_point(data=subset(accum.long_Plantes_miJuillet, labelit==T),
aes(colour=Grouping, shape=Grouping), size=2.5) +
scale_shape_manual(values = c("Graminées" = 16,
"Fleuri" = 17,
"Semé" = 15,
"Tonte récente" = 5,
"Tonte tardive" = 6)) +
geom_ribbon(aes(colour=Grouping, fill=after_scale(alpha(colour, alpha=0.15))),
show.legend=FALSE) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Nombre d'échantillonages", y = "Nombre d'espèces de plantes", colour = "Type de gestion", shape = "Type de gestion") +
theme(legend.position = "bottom",
# axis.title.x=element_blank(),
axis.title.y=element_blank())
# fin Juillet
Inv_finJuillet <- Inv_red %>%
filter(Periode == "Fin juillet")
Esp_Plant_finJuillet <- Esp_Plant %>%
rownames_to_column(var = "temp") %>%
filter(grepl("finJuillet$", temp)) %>%
column_to_rownames(var = "temp")
Accum.Plantes_finJuillet <- accumcomp(Esp_Plant_finJuillet, y=Inv_finJuillet, factor='Gestion_moment_5',
method='exact', conditioned=FALSE, plotit=FALSE)
accum.long_Plantes_finJuillet <- accumcomp.long(Accum.Plantes_finJuillet, ci=NA, label.freq=1)
accum.long_Plantes_finJuillet$Grouping <- fct_relevel(accum.long_Plantes_finJuillet$Grouping,c("Graminées", "Fleuri", "Semé", "Tonte récente", "Tonte tardive"))
pl_finJuillet <- ggplot(data=accum.long_Plantes_finJuillet, aes(x = Sites, y = Richness, ymax = UPR, ymin = LWR)) +
scale_x_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
scale_y_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
geom_line(aes(colour=Grouping), size=1.5) +
geom_point(data=subset(accum.long_Plantes_finJuillet, labelit==T),
aes(colour=Grouping, shape=Grouping), size=2.5) +
scale_shape_manual(values = c("Graminées" = 16,
"Fleuri" = 17,
"Semé" = 15,
"Tonte récente" = 5,
"Tonte tardive" = 6)) +
geom_ribbon(aes(colour=Grouping, fill=after_scale(alpha(colour, alpha=0.15))),
show.legend=FALSE) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "", y = "Nombre d'espèces de plantes", colour = "Type de gestion", shape = "Type de gestion") +
theme(legend.position = "none",
axis.title.y=element_blank())
ggarrange(pl_Juin, pl_miJuillet, pl_finJuillet, ncol = 3, common.legend = TRUE, legend="bottom") 
#ggarrange(pl_Juin, pl_miJuillet, pl_finJuillet, ncol = 1, common.legend = TRUE, legend="bottom")2.1.2.2 Pollinisateurs
# accumcomp(Esp_Poll, y = Inv_red, factor = "Gestion_2", xlim = c(1,90), plotit = T, labelit = F, rainbow = T, legend = F, xlab = "Nombre d'échantillonages", ylab = "Nombre d'espèces")[1]
# legend("topright", legend = c("Fauche", "Tonte"), col = rainbow(2), lty = 1, xpd = TRUE, inset = c(0,-0.2), horiz = F)
Accum.Poll_G2 <- accumcomp(Esp_Poll, y=Inv_red, factor='Gestion_2',
method='exact', conditioned=FALSE, plotit=FALSE)
accum.long_Poll_G2 <- accumcomp.long(Accum.Poll_G2, ci=NA, label.freq=1)
pollG2 <- ggplot(data=accum.long_Poll_G2, aes(x = Sites, y = Richness, ymax = UPR, ymin = LWR)) +
scale_x_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
scale_y_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
geom_line(aes(colour=Grouping), size=1.5) +
geom_point(data=subset(accum.long_Poll_G2, labelit==TRUE),
aes(colour=Grouping, shape=Grouping), size=2.5) +
scale_shape_manual(values = c("Fauche" = 16,
"Tonte" = 4)) +
geom_ribbon(aes(colour=Grouping, fill=after_scale(alpha(colour, alpha=0.15))),
show.legend=FALSE) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
labs(x = "Nombre d'échantillonages", y = "Nombre d'espèces de pollinisateurs", colour = "Type de gestion", shape = "Type de gestion") +
theme(legend.position = "bottom")
# accumcomp(Esp_Poll, y = Inv_red, factor = "Gestion_moment_5", col = c("blue"), xlim = c(1,70), xlab = "Nombre d'échantillonages", ylab = "Nombre d'espèces", plotit = T, labelit = F, legend = F, rainbow = F)[1]
# legend("topright", legend = c("Fleuri", "Graminées", "Semé","Tonte récente", "Tonte tardive"), col = rainbow(5), lty = 1, xpd = TRUE, inset = c(0,-0.2), horiz = F)
Accum.Poll_G5 <- accumcomp(Esp_Poll, y=Inv_red, factor='Gestion_moment_5',
method='exact', conditioned=FALSE, plotit=FALSE)
accum.long_Poll_G5 <- accumcomp.long(Accum.Poll_G5, ci=NA, label.freq=1)
accum.long_Poll_G5$Grouping <- fct_relevel(accum.long_Poll_G5$Grouping,c("Graminées", "Fleuri", "Semé", "Tonte récente", "Tonte tardive"))
pollG5 <- ggplot(data=accum.long_Poll_G5, aes(x = Sites, y = Richness, ymax = UPR, ymin = LWR)) +
scale_x_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
scale_y_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
geom_line(aes(colour=Grouping), size=1.5) +
geom_point(data=subset(accum.long_Poll_G5, labelit==T),
aes(colour=Grouping, shape=Grouping), size=2.5) +
scale_shape_manual(values = c("Graminées" = 16,
"Fleuri" = 15,
"Semé" = 17,
"Tonte récente" = 4,
"Tonte tardive" = 8)) +
geom_ribbon(aes(colour=Grouping, fill=after_scale(alpha(colour, alpha=0.15))),
show.legend=FALSE) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Nombre d'échantillonages", y = "Nombre d'espèces de pollinisateurs", colour = "Type de gestion", shape = "Type de gestion") +
theme(legend.position = "bottom")
pollG2 + pollG5
2.1.2.2.1 Périodes
# Juin
Inv_Juin <- Inv_red %>%
filter(Periode == "Juin")
Esp_Poll_Juin <- Esp_Poll %>%
rownames_to_column(var = "temp") %>%
filter(grepl("Juin$", temp)) %>%
column_to_rownames(var = "temp")
Accum.Poll_Juin <- accumcomp(Esp_Poll_Juin, y=Inv_Juin, factor='Gestion_moment_5',
method='exact', conditioned=FALSE, plotit=FALSE)
accum.long_Poll_Juin <- accumcomp.long(Accum.Poll_Juin, ci=NA, label.freq=1)
accum.long_Poll_Juin$Grouping <- fct_relevel(accum.long_Poll_Juin$Grouping,c("Graminées", "Fleuri", "Semé", "Tonte récente", "Tonte tardive"))
poll_Juin <- ggplot(data=accum.long_Poll_Juin, aes(x = Sites, y = Richness, ymax = UPR, ymin = LWR)) +
scale_x_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
scale_y_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
geom_line(aes(colour=Grouping), size=1.5) +
geom_point(data=subset(accum.long_Poll_Juin, labelit==T),
aes(colour=Grouping, shape=Grouping), size=2.5) +
scale_shape_manual(values = c("Graminées" = 16,
"Fleuri" = 17,
"Semé" = 15,
"Tonte récente" = 5,
"Tonte tardive" = 6)) +
geom_ribbon(aes(colour=Grouping, fill=after_scale(alpha(colour, alpha=0.15))),
show.legend=FALSE) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "", y = "Nombre de taxa de pollinisateurs", colour = "Type de gestion", shape = "Type de gestion") +
theme(legend.position = "none")
# mi Juillet
Inv_miJuillet <- Inv_red %>%
filter(Periode == "Mi-juillet")
Esp_Poll_miJuillet <- Esp_Poll %>%
rownames_to_column(var = "temp") %>%
filter(grepl("miJuillet$", temp)) %>%
column_to_rownames(var = "temp")
Accum.Poll_miJuillet <- accumcomp(Esp_Poll_miJuillet, y=Inv_miJuillet, factor='Gestion_moment_5',
method='exact', conditioned=FALSE, plotit=FALSE)
accum.long_Poll_miJuillet <- accumcomp.long(Accum.Poll_miJuillet, ci=NA, label.freq=1)
accum.long_Poll_miJuillet$Grouping <- fct_relevel(accum.long_Poll_miJuillet$Grouping,c("Graminées", "Fleuri", "Semé", "Tonte récente", "Tonte tardive"))
poll_miJuillet <- ggplot(data=accum.long_Poll_miJuillet, aes(x = Sites, y = Richness, ymax = UPR, ymin = LWR)) +
scale_x_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
scale_y_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
geom_line(aes(colour=Grouping), size=1.5) +
geom_point(data=subset(accum.long_Poll_miJuillet, labelit==T),
aes(colour=Grouping, shape=Grouping), size=2.5) +
scale_shape_manual(values = c("Graminées" = 16,
"Fleuri" = 17,
"Semé" = 15,
"Tonte récente" = 5,
"Tonte tardive" = 6)) +
geom_ribbon(aes(colour=Grouping, fill=after_scale(alpha(colour, alpha=0.15))),
show.legend=FALSE) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Nombre d'échantillonages", y = "Nombre de taxa de pollinisateurs", colour = "Type de gestion", shape = "Type de gestion") +
theme(legend.position = "bottom",
# axis.title.x=element_blank(),
axis.title.y=element_blank())
# fin Juillet
Inv_finJuillet <- Inv_red %>%
filter(Periode == "Fin juillet")
Esp_Poll_finJuillet <- Esp_Poll %>%
rownames_to_column(var = "temp") %>%
filter(grepl("finJuillet$", temp)) %>%
column_to_rownames(var = "temp")
Accum.Poll_finJuillet <- accumcomp(Esp_Poll_finJuillet, y=Inv_finJuillet, factor='Gestion_moment_5',
method='exact', conditioned=FALSE, plotit=FALSE)
accum.long_Poll_finJuillet <- accumcomp.long(Accum.Poll_finJuillet, ci=NA, label.freq=1)
accum.long_Poll_finJuillet$Grouping <- fct_relevel(accum.long_Poll_finJuillet$Grouping,c("Graminées", "Fleuri", "Semé", "Tonte récente", "Tonte tardive"))
poll_finJuillet <- ggplot(data=accum.long_Poll_finJuillet, aes(x = Sites, y = Richness, ymax = UPR, ymin = LWR)) +
scale_x_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
scale_y_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
geom_line(aes(colour=Grouping), size=1.5) +
geom_point(data=subset(accum.long_Poll_finJuillet, labelit==T),
aes(colour=Grouping, shape=Grouping), size=2.5) +
scale_shape_manual(values = c("Graminées" = 16,
"Fleuri" = 17,
"Semé" = 15,
"Tonte récente" = 5,
"Tonte tardive" = 6)) +
geom_ribbon(aes(colour=Grouping, fill=after_scale(alpha(colour, alpha=0.15))),
show.legend=FALSE) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "", y = "Nombre de taxa de pollinisateurs", colour = "Type de gestion", shape = "Type de gestion") +
theme(legend.position = "none",
axis.title.y=element_blank())
ggarrange(poll_Juin, poll_miJuillet, poll_finJuillet, ncol = 3, common.legend = TRUE, legend="bottom") 
#ggarrange(poll_Juin, poll_miJuillet, poll_finJuillet, ncol = 1, common.legend = TRUE, legend="bottom")2.1.3 Catégories de pollinisateurs
Inv_Classes_pl <- Inventaire %>%
rownames_to_column(var="Site_gestion_date") %>%
select(Site_gestion_date, Aglais_io:Volucella_zonaria) %>%
pivot_longer(cols = Aglais_io:Volucella_zonaria,
names_to = "Sp_poll",
values_to = "Donnees") %>%
mutate(Sp_poll = as.factor(Sp_poll))
Inv_Classes_join_pw <- left_join(Inv_Classes_pl, Classes, by = c("Sp_poll" = "Sp_Pollinisateurs")) %>%
pivot_wider(names_from = Classe_Poll,
values_from = Donnees,
values_fill = 0) %>%
select(-Sp_poll)
Inv_Classes <- aggregate(.~ Site_gestion_date, data=Inv_Classes_join_pw, FUN=sum)
Inv_noms <- Inventaire %>%
rownames_to_column(var="Site_gestion_date") %>%
select(Site, Site_gestion_date,
Nombre_quadrats,
Gestion_2, Parcelle, Gestion_4,
Mixte_isole,
Gestion_moment_4, Gestion_moment_5,
Periode,
Area_gis_m_sq,
Quartier,
Activite,
Temperature, Meteo)
Inv_fulljoin <- full_join(Site_gestion, Inv_noms, by = "Site_gestion_date") %>%
select(-Site_gestion_date_Quadrat) %>%
distinct(Site_gestion_date, .keep_all = TRUE)
Inventaire_Classes <- full_join(Inv_fulljoin, Inv_Classes, by = "Site_gestion_date")
Inventaire_Classes$S_class <- specnumber(Inventaire_Classes[16:21])
Inventaire_Classes$Ab_class <- rowSums(Inventaire_Classes[16:21])2.2 Réseaux d’interactions
2.2.1 Préparation des données et tests
Inv_noms <- Inventaire %>%
rownames_to_column(var="Site_gestion_date") %>%
select(Site, Site_gestion_date,
Nombre_quadrats,
Gestion_2, Parcelle, Gestion_4,
Mixte_isole,
Gestion_moment_4, Gestion_moment_5,
Periode,
Area_gis_m_sq,
Quartier,
Activite,
Temperature, Meteo)
Inv_fulljoin <- full_join(Site_gestion, Inv_noms, by = "Site_gestion_date")
Interactions_Gestion <- full_join(Inv_fulljoin, Interactions, by = "Site_gestion_date_Quadrat")
Interactions_Classes <- left_join(Interactions_Gestion, Classes, by = "Sp_Pollinisateurs")
Interactions_Gestion$Sp_Plantes <- sub("_", " ", Interactions_Gestion$Sp_Plantes)
Interactions_Gestion$Sp_Pollinisateurs <- sub("_", " ", Interactions_Gestion$Sp_Pollinisateurs)
Interactions_Gestion$Sp_Plantes <- sub("_", " - ", Interactions_Gestion$Sp_Plantes)
Interactions_Gestion$Sp_Pollinisateurs <- sub("_", " - ", Interactions_Gestion$Sp_Pollinisateurs)
Interactions_Gestion$Sp_Plantes <- sub("._", " ", Interactions_Gestion$Sp_Plantes)
Interactions_Gestion$Sp_Plantes <- sub("_", " ", Interactions_Gestion$Sp_Plantes)
Interactions_Gestion$Sp_Pollinisateurs <- sub("_", " ", Interactions_Gestion$Sp_Pollinisateurs)
Interactions_Classes$Sp_Plantes <- sub("_", " ", Interactions_Classes$Sp_Plantes)
Interactions_Classes$Sp_Pollinisateurs <- sub("_", " ", Interactions_Classes$Sp_Pollinisateurs)
Interactions_Classes$Sp_Plantes <- sub("_", " - ", Interactions_Classes$Sp_Plantes)
Interactions_Classes$Sp_Pollinisateurs <- sub("_", " - ", Interactions_Classes$Sp_Pollinisateurs)
Interactions_Classes$Sp_Plantes <- sub("._", " ", Interactions_Classes$Sp_Plantes)
Interactions_Classes$Sp_Plantes <- sub("_", " ", Interactions_Classes$Sp_Plantes)
Interactions_Classes$Sp_Pollinisateurs <- sub("_", " ", Interactions_Classes$Sp_Pollinisateurs)2.2.1.1 Tests
Interact_esp <- Interactions_Gestion %>%
group_by(Site_gestion_date, Sp_Plantes, Sp_Pollinisateurs) %>%
summarise(sum = sum(N_Interactions)) %>%
mutate(sum = as.numeric(sum))
network_Hemptinne_Tonte_Juin <- Interact_esp %>%
filter(Site_gestion_date == "Hemptinne_Tonte_Juin") %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
arrange(Sp_Plantes)
network_Mendel_Tonte_finJuillet <- Interact_esp %>%
filter(Site_gestion_date == "Mendel_Tonte_finJuilllet") %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
column_to_rownames(var="Sp_Plantes")
network_Reaumur_Fauche_miJuillet <- Interact_esp %>%
filter(Site_gestion_date == "Reaumur_Fauche_miJuillet") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Site_gestion_date) %>%
column_to_rownames(var="Sp_Plantes")
network_Reaumur_Fauche_miJuillet
# %>%
# kbl() %>%
# kable_classic(full_width = F, html_font = "Cambria")plotweb(network_Reaumur_Fauche_miJuillet, text.rot = 90,y.width.high=0.07, y.width.low=0.07,y.lim=c(-0.5,2))
#plotweb(network_Reaumur_Fauche_miJuillet,col.high=rainbow(15),col.low=rainbow(8),text.rot=0,y.width.high=0.07, y.width.low=0.07,y.lim=c(-0.5,2))
networklevel(network_Reaumur_Fauche_miJuillet)
# %>%
# kbl() %>%
# kable_classic(full_width = F, html_font = "Cambria") 2.2.2 Interactions
Interactions_Gestion %>% ggplot(aes (x = Gestion_2, y = N_Interactions, color = Gestion_2)) +
geom_boxplot(alpha = 0.70) +
facet_wrap(~ Periode) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
theme(legend.position = "none") +
Interactions_Gestion %>% ggplot(aes (x = Periode, y = N_Interactions, color = Periode)) +
geom_boxplot(alpha = 0.70) +
facet_wrap(~ Gestion_2) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
theme(legend.position = "none") 
Interactions_Gestion %>% ggplot(aes (x = Gestion_moment_5, y = N_Interactions, color = Gestion_moment_5)) +
geom_boxplot(alpha = 0.70) +
facet_wrap(~ Periode) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
theme(legend.position = "none") +
Interactions_Gestion %>% ggplot(aes (x = Periode, y = N_Interactions, color = Periode)) +
geom_boxplot(alpha = 0.70) +
facet_wrap(~ Gestion_moment_5) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
theme(legend.position = "none") 
Interactions_Gestion %>% ggplot(aes (x = Gestion_moment_5, y = N_Interactions, color = Periode)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
theme(legend.position = "none") +
Interactions_Gestion %>% ggplot(aes (x = Periode, y = N_Interactions, color = Gestion_moment_5)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
theme(legend.position = "none") 
Interactions_Gestion %>%
group_by(Site_gestion_date, Periode, Gestion_moment_5) %>%
summarize(n= sum(N_Interactions)) %>%
ggplot(aes (x = Gestion_moment_5, y = n, color = Gestion_moment_5)) +
geom_boxplot(alpha = 0.70) +
facet_wrap(~ Periode) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
theme(legend.position = "none") +
Interactions_Gestion %>%
group_by(Site_gestion_date, Periode, Gestion_moment_5) %>%
summarize(n= sum(N_Interactions)) %>%
ggplot(aes (x = Periode, y = n, color = Periode)) +
geom_boxplot(alpha = 0.70) +
facet_wrap(~ Gestion_moment_5) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
theme(legend.position = "none") 
Interactions_Gestion %>%
group_by(Site_gestion_date, Periode, Gestion_moment_5) %>%
summarize(n= sum(N_Interactions)) %>%
ggplot(aes (x = Periode, y = n, color = Gestion_moment_5)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
theme(legend.position = "none") +
Interactions_Gestion %>%
group_by(Site_gestion_date, Periode, Gestion_moment_5) %>%
summarize(n= sum(N_Interactions)) %>%
ggplot(aes (x = Gestion_moment_5, y = n, color = Periode)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
theme(legend.position = "none") 
Interactions_Gestion %>% ggplot(aes (x = Qtte_Plantes, y = N_Interactions, color = Gestion_moment_5)) +
geom_point() +
geom_smooth(method = "glm", se = T) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
theme(legend.position = "none") 
2.2.3 For-Loop réseaux
Loop <- Interactions_Gestion %>%
filter(Site_gestion_date != "Aula_Tonte_miJuillet",
Site_gestion_date != "AvGL_Tonte_finJuillet",
Site_gestion_date != "AvGL_Tonte_miJuillet",
Site_gestion_date != "BlancsChevaux_Tonte_finJuillet",
Site_gestion_date != "BlancsChevaux_Tonte_miJuillet",
Site_gestion_date != "Curie_Tonte_miJuillet",
Site_gestion_date != "Gare_bus_Tonte_finJuillet",
Site_gestion_date != "Lauzelle_Fauche_Juin",
Site_gestion_date != "Lauzelle_Fauche_miJuillet",
Site_gestion_date != "Lavoisier_Fauche_miJuillet",
Site_gestion_date != "Mendel_Tonte_finJuillet",
Site_gestion_date != "Mendel_Tonte_miJuillet",
Site_gestion_date != "Parc_Lapins_Fauche_Juin",
Site_gestion_date != "Parc_Lapins_Fauche_miJuillet") %>%
mutate(Site_gestion_date = as.factor(Site_gestion_date),
Sp_Plantes = as.factor(Sp_Plantes),
Sp_Pollinisateurs = as.factor(Sp_Pollinisateurs))
loop_sum <- Loop %>%
group_by(Site_gestion_date, Sp_Plantes, Sp_Pollinisateurs) %>%
summarise(sum = sum(N_Interactions))
col_names <- colnames(loop_sum)
col_names <- col_names[-1]
par(font = 3)
for (x in levels(loop_sum$Site_gestion_date)) {
pdf(paste0("Output/Reseaux_loop/Plots/", x, ".pdf")) # ouvrir le fichier
plot_nw <- plotweb(loop_sum %>%
filter(Site_gestion_date == x) %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Site_gestion_date) %>%
column_to_rownames(var="Sp_Plantes"),
high.lablength= 35, low.lablength=27,
text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3))
plot_nw # imprimer le graphique
dev.off() # fermer le fichier
}
for (x in levels(loop_sum$Site_gestion_date)) {
nw_lv <- networklevel(loop_sum %>%
filter(Site_gestion_date == x) %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Site_gestion_date) %>%
column_to_rownames(var="Sp_Plantes"))
write.csv2(nw_lv, file = paste0("Output/Reseaux_loop/Nw_Levels/", x, ".csv"))
}2.2.4 Fauche vs Tonte G2
Interact_G2 <- Interactions_Gestion %>%
group_by(Gestion_2, Sp_Plantes, Sp_Pollinisateurs) %>%
summarise(sum = sum(N_Interactions)) %>%
mutate(sum = as.numeric(sum))
par(font = 3)
network_fauche <- Interact_G2 %>%
filter(Gestion_2 == "Fauche") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Gestion_2) %>%
column_to_rownames(var="Sp_Plantes")
# plotweb(network_fauche, high.lablength= 35, low.lablength=27, low.y = -0.1, high.y = 0.7, col.high=c("#aa1e0f"),col.low=c("#aa1e0f"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,2.5))
plotweb(network_fauche, col.high=c("#aa1e0f"),col.low=c("#aa1e0f"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
#plotweb(network_fauche,col.high=c("black","yellow"),col.low=c("light green", "dark green"),text.rot=90,y.width.high=0.07, y.width.low=0.07,y.lim=c(-1,3))
#plotweb(network_fauche,col.high=rainbow(47),col.low=rainbow(47),text.rot=90,y.width.high=0.07, y.width.low=0.07,y.lim=c(-1,3))
# #plotweb(network_fauche, method = "normal",
# col.interaction=c(rep("grey50", 2),"darkgoldenrod1",rep("grey50", 2), "white"),
# col.high = c(rep("grey50", 2),"darkgoldenrod1",rep("grey50", 2), "white"),
# bor.col.high = c(rep("grey50", 2),"darkgoldenrod1",rep("grey50", 2) ,"white"),
# bor.col.interaction = c(rep("grey50", 2),"darkgoldenrod1",rep("grey50", 2), "white"),
# col.low=c("black", "white"), bor.col.low = c("black", "white"),
# text.high.col = c(rep("black", 5), "white"), text.low.col = c("black", "white"),
# text.rot = 90)
#title(main = "Réseau d'interaction fauche")
# networklevel(network_fauche, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality")) %>%
# kbl() %>%
# kable_classic(full_width = F, html_font = "Cambria")
network_tonte <- Interact_G2 %>%
filter(Gestion_2 == "Tonte") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Gestion_2) %>%
column_to_rownames(var="Sp_Plantes")
plotweb(network_tonte,col.high=c("#12661f"),col.low=c("#12661f"),text.rot=90,y.width.high=0.07, y.width.low=0.07,y.lim=c(-1,3.5))
#plotweb(network_tonte,col.high=rainbow(40),col.low=rainbow(10),text.rot=90,y.width.high=0.07, y.width.low=0.07,y.lim=c(-1,3))
# networklevel(network_tonte, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality")) %>%
# kbl() %>%
# kable_classic(full_width = F, html_font = "Cambria")
Fauche <- networklevel(network_fauche, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))
Tonte <- networklevel(network_tonte, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))
Table_NW_G2 <- cbind(Fauche, Tonte)
colnames(Table_NW_G2) <- c("Fauche", "Tonte")
Table_NW_G2## Fauche Tonte
## connectance 0.1181478 0.2943723
## web asymmetry 0.4590164 0.6500000
## weighted nestedness 0.6686565 0.4727312
## linkage density 6.9668799 2.1539505
## Fisher alpha 102.2384405 13.8299152
## Shannon diversity 4.5609652 1.6289212
## interaction evenness 0.5711813 0.2993010
## H2 0.4291655 0.6866345
## robustness.HL 0.7905750 0.8376929
## robustness.LL 0.6427648 0.5099233
## generality.HL 5.2007056 1.3405415
## vulnerability.LL 8.7330541 2.9673596# %>%
# kbl(caption = "Gestion - biclassification") %>%
# kable_classic(full_width = F, html_font = "Cambria") #%>% kable_styling() %>%
# save_kable(file = "Output/Tableau/NW_G2.pdf")2.2.5 Gestion moment 5
Interact_Gm5 <- Interactions_Gestion %>%
group_by(Gestion_moment_5, Sp_Plantes, Sp_Pollinisateurs) %>%
summarise(sum = sum(N_Interactions)) %>%
mutate(sum = as.numeric(sum))
par(font = 3)
network_gram <- Interact_Gm5 %>%
filter(Gestion_moment_5 == "Graminées") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Gestion_moment_5) %>%
column_to_rownames(var="Sp_Plantes")
plotweb(network_gram, col.high=c("#fbcb09"),col.low=c("#fbcb09"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
# networklevel(network_gram, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))
network_fleuri <- Interact_Gm5 %>%
filter(Gestion_moment_5 == "Fleuri") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Gestion_moment_5) %>%
column_to_rownames(var="Sp_Plantes")
# plotweb(network_fleuri, col.high=c("#ff7207"),col.low=c("#ff7207"),high.lablength= 30, low.lablength=35, low.y = -0.25, text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
plotweb(network_fleuri, col.high=c("#ff7207"),col.low=c("#ff7207"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
# networklevel(network_fleuri, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))
network_seme <- Interact_Gm5 %>%
filter(Gestion_moment_5 == "Semé") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Gestion_moment_5) %>%
column_to_rownames(var="Sp_Plantes")
plotweb(network_seme, col.high=c("#de1e21"),col.low=c("#de1e21"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
# networklevel(network_seme, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))
network_TonRec <- Interact_Gm5 %>%
filter(Gestion_moment_5 == "Tonte récente") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Gestion_moment_5) %>%
column_to_rownames(var="Sp_Plantes")
plotweb(network_TonRec, col.high=c("#6abe1d"), col.low=c("#6abe1d"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
# networklevel(network_TonRec, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))
network_TonTard <- Interact_Gm5 %>%
filter(Gestion_moment_5 == "Tonte tardive") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Gestion_moment_5) %>%
column_to_rownames(var="Sp_Plantes")
plotweb(network_TonTard, col.high=c("#2b790c"), col.low=c("#2b790c"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
# networklevel(network_TonTard, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))
Graminees <- networklevel(network_gram, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality")) %>%
as.data.frame()
Fleuri <- networklevel(network_fleuri, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))%>%
as.data.frame()
Seme <- networklevel(network_seme, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality")) %>%
as.data.frame()
Tonte_recente <- networklevel(network_TonRec, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))%>%
as.data.frame()
Tonte_tardive <- networklevel(network_TonTard, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality")) %>%
as.data.frame()
Table_NW_G5 <- cbind(Graminees, Fleuri, Seme, Tonte_recente, Tonte_tardive)
colnames(Table_NW_G5) <- c("Graminées", "Fleuri", "Semé", "Tonte récente", "Tonte tardive")
Table_NW_G5 ## Graminées Fleuri Semé Tonte récente
## connectance 0.2407407 0.1142534 0.1418764 0.4722222
## web asymmetry 0.6363636 0.4468085 0.4250000 0.6000000
## weighted nestedness 0.3958483 0.5646260 0.6423572 0.3757457
## linkage density 4.5620168 5.5351438 4.9818744 1.9228986
## Fisher alpha 15.7290348 71.1656509 54.1504976 4.6059563
## Shannon diversity 3.2435071 4.2667950 3.8437780 1.2878123
## interaction evenness 0.6375323 0.5706099 0.5354536 0.3593709
## H2 0.7336450 0.5114385 0.4496612 0.8104420
## robustness.HL 0.8052689 0.7730934 0.7710849 0.7025725
## robustness.LL 0.4016080 0.6190162 0.6374439 0.3201852
## generality.HL 1.4745200 3.8796971 3.3251203 1.1012144
## vulnerability.LL 7.6495137 7.1905906 6.6386285 2.7445828
## Tonte tardive
## connectance 0.2857143
## web asymmetry 0.6500000
## weighted nestedness 0.4584951
## linkage density 2.1270847
## Fisher alpha 13.6693767
## Shannon diversity 1.6404026
## interaction evenness 0.3014106
## H2 0.6990216
## robustness.HL 0.8498949
## robustness.LL 0.5047716
## generality.HL 1.3308515
## vulnerability.LL 2.9233178# %>%
# kbl(caption = "Gestion - pentaclassification") %>%
# kable_classic(full_width = F, html_font = "Cambria") 2.2.6 Période
Interact_Periode <- Interactions_Gestion %>%
group_by(Periode, Sp_Plantes, Sp_Pollinisateurs) %>%
summarise(sum = sum(N_Interactions)) %>%
mutate(sum = as.numeric(sum))
par(font = 3)
network_Juin <- Interact_Periode %>%
filter(Periode == "Juin") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Periode) %>%
column_to_rownames(var="Sp_Plantes")
plotweb(network_Juin, col.high=c("#74a9cf"), col.low=c("#74a9cf"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
# networklevel(network_Juin, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))
network_miJuil <- Interact_Periode %>%
filter(Periode == "Mi-juillet") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Periode) %>%
column_to_rownames(var="Sp_Plantes")
plotweb(network_miJuil, col.high=c("#2b8cbe"), col.low=c("#2b8cbe"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
# networklevel(network_miJuil, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))
network_finJuil <- Interact_Periode %>%
filter(Periode == "Fin juillet") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Periode) %>%
column_to_rownames(var="Sp_Plantes")
plotweb(network_finJuil, col.high=c("#045a8d"), col.low=c("#045a8d"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
# networklevel(network_finJuil, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))
juin <- networklevel(network_Juin, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))
mijuillet <- networklevel(network_miJuil, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))
finjuillet <- networklevel(network_finJuil, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))
Table_NW_P <- cbind(juin, mijuillet, finjuillet)
colnames(Table_NW_P) <- c("Juin", "Mi-juillet", "Fin juillet")
Table_NW_P ## Juin Mi-juillet Fin juillet
## connectance 0.1352785 0.1303030 0.1064815
## web asymmetry 0.3809524 0.3924051 0.3846154
## weighted nestedness 0.5246653 0.5805779 0.5296713
## linkage density 4.7004410 4.2604178 3.6859752
## Fisher alpha 61.9488092 44.7255342 40.7488484
## Shannon diversity 3.7803507 3.5367435 3.1842605
## interaction evenness 0.5165444 0.4922134 0.4442924
## H2 0.5157954 0.4665388 0.5867317
## robustness.HL 0.7788051 0.7668089 0.7323744
## robustness.LL 0.6363928 0.6371332 0.5838619
## generality.HL 3.9909709 3.5874657 3.0955332
## vulnerability.LL 5.4099111 4.9333699 4.2764171# %>%
# kbl(caption = "Période") %>%
# kable_classic(full_width = F, html_font = "Cambria") 2.2.7 Période x Gestion moment 2
Interact_Periode_g2 <- Interactions_Gestion %>%
group_by(Periode, Gestion_2, Sp_Plantes, Sp_Pollinisateurs) %>%
summarise(sum = sum(N_Interactions)) %>%
mutate(sum = as.numeric(sum))
par(font = 3)
network_Juin_fauche <- Interact_Periode_g2 %>%
filter(Periode == "Juin" &
Gestion_2 == "Fauche") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Periode) %>%
select(-Gestion_2) %>%
column_to_rownames(var="Sp_Plantes")
plotweb(network_Juin_fauche, col.high=c("#74a9cf","#aa1e0f"), col.low=c("#74a9cf","#aa1e0f"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3))
networklevel(network_Juin_fauche, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))## connectance web asymmetry weighted nestedness
## 0.1272727 0.3580247 0.5479423
## linkage density Fisher alpha Shannon diversity
## 5.9304150 68.8198247 4.4090059
## interaction evenness H2 robustness.HL
## 0.6068472 0.4613997 0.7640749
## robustness.LL generality.HL vulnerability.LL
## 0.6321896 4.6183087 7.2425214network_Juin_tonte <- Interact_Periode_g2 %>%
filter(Periode == "Juin" &
Gestion_2 == "Tonte") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Periode) %>%
select(-Gestion_2) %>%
column_to_rownames(var="Sp_Plantes")
plotweb(network_Juin_tonte, col.high=c("#74a9cf","#12661f"), col.low=c("#74a9cf","#12661f"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3))
networklevel(network_Juin_tonte, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))## connectance web asymmetry weighted nestedness
## 0.3115942 0.5862069 0.4124064
## linkage density Fisher alpha Shannon diversity
## 2.0355532 10.0881095 1.7214311
## interaction evenness H2 robustness.HL
## 0.3493693 0.7407693 0.7764011
## robustness.LL generality.HL vulnerability.LL
## 0.4487474 1.3961494 2.6749570network_miJui_fauche <- Interact_Periode_g2 %>%
filter(Periode == "Mi-juillet" &
Gestion_2 == "Fauche") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Periode) %>%
select(-Gestion_2) %>%
column_to_rownames(var="Sp_Plantes")
plotweb(network_miJui_fauche, col.high=c("#2b8cbe","#aa1e0f"), col.low=c("#2b8cbe","#aa1e0f"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3))
networklevel(network_miJui_fauche, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))## connectance web asymmetry weighted nestedness
## 0.1279264 0.3866667 0.5894706
## linkage density Fisher alpha Shannon diversity
## 4.5996354 43.6299928 3.7923862
## interaction evenness H2 robustness.HL
## 0.5351385 0.5190806 0.7505902
## robustness.LL generality.HL vulnerability.LL
## 0.6165338 3.6206267 5.5786441network_miJui_tonte <- Interact_Periode_g2 %>%
filter(Periode == "Mi-juillet" &
Gestion_2 == "Tonte") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Periode) %>%
select(-Gestion_2) %>%
column_to_rownames(var="Sp_Plantes")
plotweb(network_miJui_tonte, col.high=c("#2b8cbe","#12661f"), col.low=c("#2b8cbe","#12661f"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3))
networklevel(network_miJui_tonte, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))## connectance web asymmetry weighted nestedness
## 0.3000000 0.5652174 0.5032866
## linkage density Fisher alpha Shannon diversity
## 2.1902785 5.7177643 1.4246274
## interaction evenness H2 robustness.HL
## 0.3165973 0.7622993 0.7245617
## robustness.LL generality.HL vulnerability.LL
## 0.3952795 1.1260362 3.2545208network_finJui_fauche <- Interact_Periode_g2 %>%
filter(Periode == "Fin juillet" &
Gestion_2 == "Fauche") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Periode) %>%
select(-Gestion_2) %>%
column_to_rownames(var="Sp_Plantes")
plotweb(network_finJui_fauche, col.high=c("#045a8d","#aa1e0f"), col.low=c("#045a8d","#aa1e0f"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3))
networklevel(network_finJui_fauche, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))## connectance web asymmetry weighted nestedness
## 0.1072727 0.3888889 0.5493924
## linkage density Fisher alpha Shannon diversity
## 4.7715399 42.7795835 3.9215302
## interaction evenness H2 robustness.HL
## 0.5599734 0.5423388 0.7005818
## robustness.LL generality.HL vulnerability.LL
## 0.5953805 3.1904334 6.3526464network_finJui_tonte <- Interact_Periode_g2 %>%
filter(Periode == "Fin juillet" &
Gestion_2 == "Tonte") %>%
filter(sum!=0) %>%
select(Sp_Pollinisateurs,Sp_Plantes,sum) %>%
arrange(Sp_Pollinisateurs) %>%
pivot_wider(names_from = Sp_Pollinisateurs,
values_from = sum,
values_fill = 0) %>%
select(where(~ any(. != 0))) %>%
arrange(Sp_Plantes) %>%
as.data.frame() %>%
select(-Periode) %>%
select(-Gestion_2) %>%
column_to_rownames(var="Sp_Plantes")
plotweb(network_finJui_tonte, col.high=c("#045a8d","#12661f"), col.low=c("#045a8d","#12661f"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3))
networklevel(network_finJui_tonte, index=c("connectance", "web asymmetry", "weighted nestedness", "linkage density", "Fisher alpha","Shannon diversity", "interaction evenness","H2","robustness", "robustness", "generality"))## connectance web asymmetry weighted nestedness
## 0.2368421 0.5200000 0.4073270
## linkage density Fisher alpha Shannon diversity
## 1.4223746 6.0066788 0.9281065
## interaction evenness H2 robustness.HL
## 0.1959602 0.7960476 0.7000583
## robustness.LL generality.HL vulnerability.LL
## 0.4198334 1.2289000 1.61584912.3 Expérience supplémentaire: Tonte
#Mise en facteur des variables
expe_Tonte = expe_Tonte %>%
mutate(Site = as.factor(Site),
Date = as.factor(Date))
#Tonte$Date <- dmy(Tonte$Date)
summary(expe_Tonte)
expe_Tonte$S <- specnumber(expe_Tonte[5:11])
expe_Tonte$Ab <- rowSums(expe_Tonte[5:11])
# mettre une ligne par espèce
Tonte_pl <- expe_Tonte %>%
pivot_longer(cols = Bellis_perennis:Trifolium_repens,
names_to = "Espèces",
values_to = "Qtté")
Tonte_pl$Qtte_trans <- Tonte_pl$Qtté
Tonte_pl[Tonte_pl == 0] <- NAexpe_Tonte %>% ggplot(aes (x = Jour_af_tonte, y = Ab, color=Site)) +
geom_line() +
labs(title = "Abondance de fleurs après la tonte selon les sites" , x = "Jour après la tonte", y = "Abondance de fleurs") +
theme(legend.position="bottom") +
expe_Tonte %>% ggplot(aes (x = Jour_af_tonte, y = Ab, color=Site)) +
geom_smooth()+
labs(title = "Abondance de fleurs après la tonte selon les sites" , x = "Jour après la tonte", y = "Abondance de fleurs") +
theme(legend.position="none")
Tonte_pl %>% ggplot(aes (x = Jour_af_tonte, y = S)) +
facet_wrap(~Site) +
geom_smooth(se = F)+
labs(title = "Abondance de fleurs après la tonte selon les sites" , x = "Jour après la tonte", y = "Abondance de fleurs") +
theme(legend.position="bottom") +
Tonte_pl %>% ggplot(aes (x = Jour_af_tonte, y = S, color = Site)) +
geom_point() +
geom_smooth(se = F)+
labs(title = "Richesse spécifique en fleurs après la tonte selon les sites" , x = "Jour après la tonte", y = "Richesse spécifique en fleurs") +
theme(legend.position="bottom")
Tonte_pl %>% ggplot(aes (x = Jour_af_tonte, y = Qtté, color=Espèces)) +
geom_smooth(se=F)+
labs(title = "Abondance de fleurs après la tonte selon les sites" , x = "Jour après la tonte", y = "Abondance de fleurs") +
theme(legend.position="bottom") 
Tonte_pl %>% ggplot(aes (x = Jour_af_tonte, y = Qtté, color=Espèces)) +
facet_wrap(~Site) +
geom_smooth(se = F)+
labs(title = "Abondance de fleurs après la tonte selon les sites" , x = "Jour après la tonte", y = "Abondance de fleurs") +
theme(legend.position="bottom")
3 Statistiques
3.1 Multivariée
3.1.1 ACP
3.1.1.1 ACP - Inv
gestion_2 <- as.factor(Inv$Gestion_2)
gestion_5 <- as.factor(Inv$Gestion_moment_5)
periode <- as.factor(Inv$Periode)
activite <- as.factor(Inv$Activite)
meteo <- as.factor(Inv$Meteo)
site <- as.factor(Inv$Site)
quartier <- as.factor(Inv$Quartier)
NumVar <- sapply(Inv, is.numeric)
df <- Inv[,NumVar] # %>%
# select(!c(Heure_fin,Heure_debut))
n <- nrow(df)
p <- ncol(df)
acp <- dudi.pca(df, scannf = F, nf = p)
fviz_pca_biplot(acp)
screeplot(acp, main = "Valeurs propres")
fviz_pca_var(acp,
col.var = "cos2",
title = "Cercle de corrélation et qualité de représentation dans le plan 1-2",
gradient.cols=c("red", "gold", "forestgreen")) 
fviz_pca_var(acp,
col.var = "cos2", axes = c(2,3),
title = "Cercle de corrélation et qualité de représentation dans le plan 2-3",
gradient.cols=c("red", "gold", "forestgreen")) 
fviz_pca_var(acp,
col.var = "cos2", axes = c(4,5),
title = "Cercle de corrélation et qualité de représentation dans le plan 4-5",
gradient.cols=c("red", "gold", "forestgreen")) 
fviz_pca_var(acp,
col.var = "cos2", axes = c(5,6),
title = "Cercle de corrélation et qualité de représentation dans le plan 5-6",
gradient.cols=c("red", "gold", "forestgreen"))
fviz_pca_ind(acp, repel = T,
labelsize = 2,
col.ind = "cos2",
gradient.cols=c("red", "gold", "forestgreen"),
title = "Qualité de représentation des individus dans le plan 1-2")
fviz_pca_ind(acp, axes = c(2,3),
labelsize = 0,
col.ind = "cos2",
gradient.cols=c("red", "gold", "forestgreen"),
title = "Qualité de représentation des individus dans le plan 2-3")
fviz_pca_ind(acp,
labelsize = 0,
col.ind = gestion_2,
addEllipses = T,
title = "Position des individus dans le plan 1-2 selon la variable gestion_2") +
aes(color = gestion_2, shape = gestion_2, fill = gestion_2) +
scale_shape_manual(values = c("Fauche" = 16,
"Tonte" = 5)) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
scale_fill_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) 
fviz_pca_ind(acp,
labelsize = 0,
col.ind = gestion_5,
addEllipses = T,
title = "Position des individus dans le plan 1-2 selon la variable gestion_5") +
aes(color = gestion_5, shape = gestion_5, fill = gestion_5)+
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
scale_fill_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
scale_shape_manual(values = c("Graminées" = 16,
"Fleuri" = 17,
"Semé" = 15,
"Tonte récente" = 5,
"Tonte tardive" = 6)) 
fviz_pca_ind(acp,
labelsize = 0,
col.ind = periode,
addEllipses = T,
title = "Position des individus dans le plan 1-2 selon la variable periode") +
aes(color = periode, shape = periode, fill = periode) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
scale_fill_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d"))
fviz_pca_ind(acp, col.var = "var",
labelsize = 0,
col.ind = activite,
addEllipses = T,
title = "Position des individus dans le plan 1-2 selon la variable activite") +
aes(color = activite, shape = activite, fill = activite) +
scale_color_manual(values = c("Forte" = "#3f007d",
"Moyenne" = "#6a51a3",
"Nulle" = "#9e9ac8")) +
scale_fill_manual(values = c("Forte" = "#3f007d",
"Moyenne" = "#6a51a3",
"Nulle" = "#9e9ac8")) 
fviz_pca_ind(acp,
labelsize = 0,
col.ind = meteo,
addEllipses = T,
title = "Position des individus dans le plan 1-2 selon la variable meteo") +
aes(color = meteo, shape = meteo, fill = meteo) +
scale_color_manual(values = c("Nuageux" = "#7f908c",
"Alternances" = "#79ccbd",
"Soleil" = "#fbcb09")) +
scale_fill_manual(values = c("Nuageux" = "#7f908c",
"Alternances" = "#79ccbd",
"Soleil" = "#fbcb09")) 
fviz_pca_ind(acp,
labelsize = 0,
col.ind = quartier,
addEllipses = T,
title = "Position des individus dans le plan 1-2 selon la variable quartier") 
summary(acp)## Class: pca dudi
## Call: dudi.pca(df = df, scannf = F, nf = p)
##
## Total inertia: 7
##
## Eigenvalues:
## Ax1 Ax2 Ax3 Ax4 Ax5
## 3.4909 1.2737 0.7696 0.6520 0.4721
##
## Projected inertia (%):
## Ax1 Ax2 Ax3 Ax4 Ax5
## 49.871 18.195 10.995 9.314 6.745
##
## Cumulative projected inertia (%):
## Ax1 Ax1:2 Ax1:3 Ax1:4 Ax1:5
## 49.87 68.07 79.06 88.37 95.12
##
## (Only 5 dimensions (out of 7) are shown)A <- fviz_pca_ind(acp,
labelsize = 0,
col.ind = gestion_2,
addEllipses = T,
title = "Position des parcelles dans le plan 1-2 selon la variable gestion_2") +
aes(color = gestion_2, shape = gestion_2, fill = gestion_2) +
scale_shape_manual(values = c("Fauche" = 16,
"Tonte" = 5)) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
scale_fill_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
labs(color = "Type de gestion", fill = "Type de gestion", shape = "Type de gestion")
B <- fviz_pca_ind(acp,
labelsize = 0,
col.ind = gestion_5,
addEllipses = T,
title = "Position des parcelles dans le plan 1-2 selon la variable gestion_5") +
aes(color = gestion_5, shape = gestion_5, fill = gestion_5)+
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
scale_fill_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
scale_shape_manual(values = c("Graminées" = 16,
"Fleuri" = 17,
"Semé" = 15,
"Tonte récente" = 5,
"Tonte tardive" = 6)) +
labs(color = "Type de gestion", fill = "Type de gestion", shape = "Type de gestion")
C <- fviz_pca_ind(acp,
labelsize = 0,
col.ind = periode,
addEllipses = T,
title = "Position des parcelles dans le plan 1-2 selon la variable periode") +
aes(color = periode, shape = periode, fill = periode) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
scale_fill_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(color = "Période", fill = "Période", shape = "Période")
D <- fviz_pca_var(acp,
col.var = "cos2",
title = "Cercle de corrélation et qualité de représentation dans le plan 1-2",
gradient.cols=c("red", "gold", "forestgreen"))
(D+B)/(C+A) 
A <- fviz_pca_ind(acp,
labelsize = 0,
col.ind = quartier,
addEllipses = T,
palette = terrain.colors(10),
title = "Position des parcelles dans le plan 1-2 selon la variable quartier") +
labs(color = "Quartier", fill = "Quartier", shape = "Quartier")
B <- fviz_pca_ind(acp,
labelsize = 0,
col.ind = activite,
addEllipses = T,
#palette = terrain.colors(5),
title = "Position des parcelles dans le plan 1-2 selon la variable activite") +
aes(color = activite, shape = activite, fill = activite) +
scale_color_manual(values = c("Forte" = "#3f007d",
"Moyenne" = "#6a51a3",
"Nulle" = "#9e9ac8")) +
scale_fill_manual(values = c("Forte" = "#3f007d",
"Moyenne" = "#6a51a3",
"Nulle" = "#9e9ac8")) +
labs(color = "Activité", fill = "Activité", shape = "Activité")
C <- fviz_pca_ind(acp,
labelsize = 0,
col.ind = meteo,
addEllipses = T,
#palette = terrain.colors(5),
title = "Position des parcelles dans le plan 1-2 selon la variable meteo") +
aes(color = meteo, shape = meteo, fill = meteo) +
scale_color_manual(values = c("Nuageux" = "#7f908c",
"Alternances" = "#79ccbd",
"Soleil" = "#fbcb09")) +
scale_fill_manual(values = c("Nuageux" = "#7f908c",
"Alternances" = "#79ccbd",
"Soleil" = "#fbcb09")) +
labs(color = "Météo", fill = "Météo", shape = "Météo")
D <- fviz_pca_var(acp,
col.var = "cos2",
title = "Cercle de corrélation et qualité de représentation dans le plan 1-2",
gradient.cols=c("red", "gold", "forestgreen"))
(D+B)/(C+A) 
3.1.1.2 ACP - Classes
Inventaire_Classes_acp <- Inventaire_Classes %>%
column_to_rownames(var="Site_gestion_date")
NumVar <- sapply(Inventaire_Classes_acp, is.numeric)
df <- Inventaire_Classes_acp[,NumVar] %>%
select(-S_class, -Ab_class)
n <- nrow(df)
p <- ncol(df)
gestion_2 <- as.factor(Inventaire_Classes_acp$Gestion_2)
gestion_5 <- as.factor(Inventaire_Classes_acp$Gestion_moment_5)
periode <- as.factor(Inventaire_Classes_acp$Periode)
activite <- as.factor(Inventaire_Classes_acp$Activite)
meteo <- as.factor(Inventaire_Classes_acp$Meteo)
site <- as.factor(Inventaire_Classes_acp$Site)
quartier <- as.factor(Inventaire_Classes_acp$Quartier)
acp <- dudi.pca(df, scannf = F, nf = p)
fviz_pca_biplot(acp)
screeplot(acp, main = "Valeurs propres")
fviz_pca_var(acp,
col.var = "cos2",
title = "Cercle de corrélation et qualité de représentation dans le plan 1-2",
gradient.cols=c("red", "gold", "forestgreen")) 
fviz_pca_var(acp,
col.var = "cos2", axes = c(2,3),
title = "Cercle de corrélation et qualité de représentation dans le plan 2-3",
gradient.cols=c("red", "gold", "forestgreen")) 
fviz_pca_ind(acp, repel = T,
labelsize = 2,
col.ind = "cos2",
gradient.cols=c("red", "gold", "forestgreen"),
title = "Qualité de représentation des individus dans le plan 1-2")
fviz_pca_ind(acp, axes = c(2,3),
labelsize = 0,
col.ind = "cos2",
gradient.cols=c("red", "gold", "forestgreen"),
title = "Qualité de représentation des individus dans le plan 2-3")
fviz_pca_ind(acp,
labelsize = 0,
col.ind = gestion_2,
addEllipses = T,
title = "Position des individus dans le plan 1-2 selon la variable gestion_2") +
aes(color = gestion_2, shape = gestion_2, fill = gestion_2) +
scale_shape_manual(values = c("Fauche" = 16,
"Tonte" = 5)) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
scale_fill_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) 
fviz_pca_ind(acp,
labelsize = 0,
col.ind = gestion_5,
addEllipses = T,
title = "Position des individus dans le plan 1-2 selon la variable gestion_5") +
aes(color = gestion_5, shape = gestion_5, fill = gestion_5) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
scale_fill_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
scale_shape_manual(values = c("Graminées" = 16,
"Fleuri" = 17,
"Semé" = 15,
"Tonte récente" = 5,
"Tonte tardive" = 6)) 
fviz_pca_ind(acp,
labelsize = 0,
col.ind = periode,
addEllipses = T,
title = "Position des individus dans le plan 1-2 selon la variable periode") +
aes(color = periode, shape = periode, fill = periode) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
scale_fill_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d"))
fviz_pca_ind(acp, col.var = "var",
labelsize = 0,
col.ind = activite,
addEllipses = T,
title = "Position des individus dans le plan 1-2 selon la variable activite") +
aes(color = activite, shape = activite, fill = activite) +
scale_color_manual(values = c("Forte" = "#3f007d",
"Moyenne" = "#6a51a3",
"Nulle" = "#9e9ac8")) +
scale_fill_manual(values = c("Forte" = "#3f007d",
"Moyenne" = "#6a51a3",
"Nulle" = "#9e9ac8")) 
fviz_pca_ind(acp,
labelsize = 0,
col.ind = meteo,
addEllipses = T,
title = "Position des individus dans le plan 1-2 selon la variable meteo") +
aes(color = meteo, shape = meteo, fill = meteo) +
scale_color_manual(values = c("Nuageux" = "#7f908c",
"Alternances" = "#79ccbd",
"Soleil" = "#fbcb09")) +
scale_fill_manual(values = c("Nuageux" = "#7f908c",
"Alternances" = "#79ccbd",
"Soleil" = "#fbcb09")) 
fviz_pca_ind(acp,
labelsize = 0,
col.ind = quartier,
addEllipses = T,
title = "Position des individus dans le plan 1-2 selon la variable quartier") 
summary(acp)## Class: pca dudi
## Call: dudi.pca(df = df, scannf = F, nf = p)
##
## Total inertia: 9
##
## Eigenvalues:
## Ax1 Ax2 Ax3 Ax4 Ax5
## 2.6816 1.2850 1.2256 0.9067 0.8485
##
## Projected inertia (%):
## Ax1 Ax2 Ax3 Ax4 Ax5
## 29.796 14.278 13.618 10.075 9.428
##
## Cumulative projected inertia (%):
## Ax1 Ax1:2 Ax1:3 Ax1:4 Ax1:5
## 29.80 44.07 57.69 67.77 77.19
##
## (Only 5 dimensions (out of 9) are shown)A <- fviz_pca_ind(acp,
labelsize = 0,
col.ind = gestion_2,
addEllipses = T,
title = "Position des parcelles dans le plan 1-2 selon la variable gestion_2") +
aes(color = gestion_2, shape = gestion_2, fill = gestion_2) +
scale_shape_manual(values = c("Fauche" = 16,
"Tonte" = 5)) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
scale_fill_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
labs(color = "Type de gestion", fill = "Type de gestion", shape = "Type de gestion")
B <- fviz_pca_ind(acp,
labelsize = 0,
col.ind = gestion_5,
addEllipses = T,
title = "Position des parcelles dans le plan 1-2 selon la variable gestion_5") +
aes(color = gestion_5, shape = gestion_5, fill = gestion_5)+
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
scale_fill_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
scale_shape_manual(values = c("Graminées" = 16,
"Fleuri" = 17,
"Semé" = 15,
"Tonte récente" = 5,
"Tonte tardive" = 6)) +
labs(color = "Type de gestion", fill = "Type de gestion", shape = "Type de gestion")
C <- fviz_pca_ind(acp,
labelsize = 0,
col.ind = periode,
addEllipses = T,
title = "Position des parcelles dans le plan 1-2 selon la variable periode") +
aes(color = periode, shape = periode, fill = periode) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
scale_fill_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(color = "Période", fill = "Période", shape = "Période")
D <- fviz_pca_var(acp,
col.var = "cos2",
title = "Cercle de corrélation et qualité de représentation dans le plan 1-2",
gradient.cols=c("red", "gold", "forestgreen"))
(D+B)/(C+A) 
A <- fviz_pca_ind(acp,
labelsize = 0,
col.ind = quartier,
addEllipses = T,
palette = terrain.colors(10),
title = "Position des parcelles dans le plan 1-2 selon la variable quartier") +
labs(color = "Quartier", fill = "Quartier", shape = "Quartier")
B <- fviz_pca_ind(acp,
labelsize = 0,
col.ind = activite,
addEllipses = T,
title = "Position des parcelles dans le plan 1-2 selon la variable activite") +
aes(color = activite, shape = activite, fill = activite) +
scale_color_manual(values = c("Forte" = "#3f007d",
"Moyenne" = "#6a51a3",
"Nulle" = "#9e9ac8")) +
scale_fill_manual(values = c("Forte" = "#3f007d",
"Moyenne" = "#6a51a3",
"Nulle" = "#9e9ac8")) +
labs(color = "Activité", fill = "Activité", shape = "Activité")
C <- fviz_pca_ind(acp,
labelsize = 0,
col.ind = meteo,
addEllipses = T,
title = "Position des parcelles dans le plan 1-2 selon la variable meteo") +
aes(color = meteo, shape = meteo, fill = meteo) +
scale_color_manual(values = c("Nuageux" = "#7f908c",
"Alternances" = "#79ccbd",
"Soleil" = "#fbcb09")) +
scale_fill_manual(values = c("Nuageux" = "#7f908c",
"Alternances" = "#79ccbd",
"Soleil" = "#fbcb09")) +
labs(color = "Météo", fill = "Météo", shape = "Météo")
D <- fviz_pca_var(acp,
col.var = "cos2",
title = "Cercle de corrélation et qualité de représentation dans le plan 1-2",
gradient.cols=c("red", "gold", "forestgreen"))
(D+B)/(C+A) 
3.1.1.3 ACP - Fin juillet
AllFinJuillet_acp<- AllFinJuillet[,c(4,19:68)]
AllFinJuillet_acp <- AllFinJuillet_acp %>%
column_to_rownames(var="Parcelle")
NumVar <- sapply(AllFinJuillet_acp, is.numeric)
df <- AllFinJuillet_acp[,NumVar]
n <- nrow(df)
p <- ncol(df)
gestion_2 <- as.factor(AllFinJuillet$Gestion_2)
gestion_5 <- as.factor(AllFinJuillet$Gestion_moment_5)
activite <- as.factor(AllFinJuillet$Activite)
meteo <- as.factor(AllFinJuillet$Meteo)
site <- as.factor(AllFinJuillet$Site)
quartier <- as.factor(AllFinJuillet$Quartier)
acp <- dudi.pca(df, scannf = F, nf = p)
fviz_pca_biplot(acp)
screeplot(acp, main = "Valeurs propres")
fviz_pca_var(acp,
col.var = "cos2",
title = "Cercle de corrélation et qualité de représentation dans le plan 1-2",
gradient.cols=c("red", "gold", "forestgreen")) 
fviz_pca_var(acp,
col.var = "cos2", axes = c(2,3),
title = "Cercle de corrélation et qualité de représentation dans le plan 2-3",
gradient.cols=c("red", "gold", "forestgreen")) 
fviz_pca_ind(acp, repel = T,
labelsize = 2,
col.ind = "cos2",
gradient.cols=c("red", "gold", "forestgreen"),
title = "Qualité de représentation des individus dans le plan 1-2")
fviz_pca_ind(acp, axes = c(2,3),
labelsize = 0,
col.ind = "cos2",
gradient.cols=c("red", "gold", "forestgreen"),
title = "Qualité de représentation des individus dans le plan 2-3")
fviz_pca_ind(acp,
labelsize = 0,
col.ind = gestion_2,
addEllipses = T,
title = "Position des individus dans le plan 1-2 selon la variable gestion_2") +
aes(color = gestion_2, shape = gestion_2, fill = gestion_2) +
scale_shape_manual(values = c("Fauche" = 16,
"Tonte" = 5)) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
scale_fill_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) 
fviz_pca_ind(acp,
labelsize = 0,
col.ind = gestion_5,
addEllipses = T,
title = "Position des individus dans le plan 1-2 selon la variable gestion_5") +
aes(color = gestion_5, shape = gestion_5, fill = gestion_5) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
scale_fill_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
scale_shape_manual(values = c("Graminées" = 16,
"Fleuri" = 17,
"Semé" = 15,
"Tonte récente" = 5,
"Tonte tardive" = 6)) 
fviz_pca_ind(acp, col.var = "var",
labelsize = 0,
col.ind = activite,
addEllipses = T,
title = "Position des parcelles dans le plan 1-2 selon la variable activite") +
aes(color = activite, shape = activite, fill = activite) +
scale_color_manual(values = c("Forte" = "#3f007d",
"Moyenne" = "#6a51a3",
"Nulle" = "#9e9ac8")) +
scale_fill_manual(values = c("Forte" = "#3f007d",
"Moyenne" = "#6a51a3",
"Nulle" = "#9e9ac8")) 
fviz_pca_ind(acp,
labelsize = 0,
col.ind = meteo,
addEllipses = T,
title = "Position des parcelles dans le plan 1-2 selon la variable meteo") +
aes(color = meteo, shape = meteo, fill = meteo) +
scale_color_manual(values = c("Nuageux" = "#7f908c",
"Alternances" = "#79ccbd",
"Soleil" = "#fbcb09")) +
scale_fill_manual(values = c("Nuageux" = "#7f908c",
"Alternances" = "#79ccbd",
"Soleil" = "#fbcb09")) 
fviz_pca_ind(acp,
labelsize = 0,
col.ind = quartier,
addEllipses = T,
title = "Position des parcelles dans le plan 1-2 selon la variable quartier") 
summary(acp)## Class: pca dudi
## Call: dudi.pca(df = df, scannf = F, nf = p)
##
## Total inertia: 43
##
## Eigenvalues:
## Ax1 Ax2 Ax3 Ax4 Ax5
## 8.348 4.048 3.627 3.039 2.787
##
## Projected inertia (%):
## Ax1 Ax2 Ax3 Ax4 Ax5
## 19.413 9.415 8.436 7.067 6.481
##
## Cumulative projected inertia (%):
## Ax1 Ax1:2 Ax1:3 Ax1:4 Ax1:5
## 19.41 28.83 37.26 44.33 50.81
##
## (Only 5 dimensions (out of 30) are shown)A <- fviz_pca_ind(acp,
labelsize = 0,
col.ind = gestion_2,
addEllipses = T,
title = "Position des parcelles dans le plan 1-2 selon la variable gestion_2") +
aes(color = gestion_2, shape = gestion_2, fill = gestion_2) +
scale_shape_manual(values = c("Fauche" = 16,
"Tonte" = 5)) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
scale_fill_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
labs(color = "Type de gestion", fill = "Type de gestion", shape = "Type de gestion")
B <- fviz_pca_ind(acp,
labelsize = 0,
col.ind = gestion_5,
addEllipses = T,
title = "Position des parcelles dans le plan 1-2 selon la variable gestion_5") +
aes(color = gestion_5, shape = gestion_5, fill = gestion_5)+
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
scale_fill_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
scale_shape_manual(values = c("Graminées" = 16,
"Fleuri" = 17,
"Semé" = 15,
"Tonte récente" = 5,
"Tonte tardive" = 6)) +
labs(color = "Type de gestion", fill = "Type de gestion", shape = "Type de gestion")
C <- fviz_pca_ind(acp,
labelsize = 0,
col.ind = meteo,
addEllipses = T,
title = "Position des parcelles dans le plan 1-2 selon la variable meteo") +
aes(color = meteo, shape = meteo, fill = meteo) +
scale_color_manual(values = c("Nuageux" = "#7f908c",
"Alternances" = "#79ccbd",
"Soleil" = "#fbcb09")) +
scale_fill_manual(values = c("Nuageux" = "#7f908c",
"Alternances" = "#79ccbd",
"Soleil" = "#fbcb09")) +
labs(color = "Météo", fill = "Météo", shape = "Météo")
D <- fviz_pca_var(acp,
col.var = "cos2",
title = "Cercle de corrélation et qualité de représentation dans le plan 1-2",
gradient.cols=c("red", "gold", "forestgreen"))
(D+B)/(C+A) 
3.1.2 ACC
3.1.2.1 ACC - Sp plantes
Y <- Esp_Plant
X <- Inv %>%
select(!c(Site, Parcelle,Gestion_2, Gestion_4, Gestion_moment_4, Date, Heure_debut, Heure_fin, Quartier, Meteo, Jours, Mixte_isole, Activite, Nombre_quadrats))
#dim(X) ; dim(Y)
X<-X[rowSums(Y)!=0,]
Y<-Y[rowSums(Y)!=0,]
Y<-Y[,colSums(Y)!=0]
n <- nrow(Y)
#n == nrow(X)
p <- ncol(Y)
m <- ncol(X)
gestion_2 <- as.factor(X$Gestion_2)
gestion_5 <- as.factor(X$Gestion_moment_5)
afc_Y <- dudi.coa(Y, scannf = FALSE, nf = p)
acc <- pcaiv(afc_Y, X, scannf = FALSE, nf = m)
Q <- viz_coaiv(acc,"co") +
scale_x_continuous(limits=c(-1.5, 2))
viz_coaiv(acc,"li")
viz_coaiv(acc, "li", axes = c(2,3)) 
Z <- viz_coaiv(acc, "ls") +
scale_x_continuous(limits=c(-1.1, 1.9))
viz_coaiv(acc, "match")
viz_coaiv(acc, "fa") 
E <- viz_coaiv(acc, "cor") +
scale_x_continuous(limits=c(-1.5, 1.5))
viz_coaiv(acc, "cor", axes=c(1,3)) 
viz_coaiv(acc, "li", axes = c(1,3))
R <- viz_coaiv(acc, "as")
viz_coaiv(acc, "as", axes=c(2,3)) 
Z = Z + aes(color = gestion_5) + labs(title = "Position des parcelles", color = "Type de gestion", fill = "Type de gestion", shape = "Type de gestion") +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
theme (legend.position = "bottom",
text = element_text(size=12))
Q = Q + labs(title = "Position des espèces de plantes") +
theme (legend.position = "bottom",
text = element_text(size=12))
E= E + labs(title = "Variables explicatives") +
theme (legend.position = "bottom",
text = element_text(size=12))
R= R + labs(title = "Participations des axes à la dimention 1-2")
(Q + E) / (Z + R)
(Q + E) / (Z)
3.1.2.2 ACC - Sp poll
Y <- Esp_Poll
X <- Inv %>%
select(!c(Site, Parcelle,Gestion_2, Gestion_4, Gestion_moment_4, Date, Heure_debut, Heure_fin, Quartier, Meteo, Jours, Mixte_isole, Activite, Nombre_quadrats))
#dim(X) ; dim(Y)
X<-X[rowSums(Y)!=0,]
Y<-Y[rowSums(Y)!=0,]
Y<-Y[,colSums(Y)!=0]
n <- nrow(Y)
#n == nrow(X)
p <- ncol(Y)
m <- ncol(X)
gestion_2 <- as.factor(X$Gestion_2)
gestion_5 <- as.factor(X$Gestion_moment_5)
afc_Y <- dudi.coa(Y, scannf = FALSE, nf = p)
acc <- pcaiv(afc_Y, X, scannf = FALSE, nf = m)
Q <- viz_coaiv(acc,"co")
viz_coaiv(acc,"li")
viz_coaiv(acc, "li", axes = c(2,3))
Z <- viz_coaiv(acc, "ls")
viz_coaiv(acc, "match")
viz_coaiv(acc, "fa") 
E <- viz_coaiv(acc, "cor")
viz_coaiv(acc, "cor", axes=c(1,3)) 
viz_coaiv(acc, "li", axes = c(1,3))
R <- viz_coaiv(acc, "as")
viz_coaiv(acc, "as", axes=c(2,3)) 
Z = Z + aes(color = gestion_5) + labs(title = "Position des parcelles", color = "Type de gestion", fill = "Type de gestion", shape = "Type de gestion") +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c"))
Q = Q + labs(title = "Position des espèces de pollinisateurs")
E= E + labs(title = "Variables explicatives")
R= R + labs(title = "Participations des axes à la dimention 1-2")
(Q + E) / (Z + R)
3.1.2.3 ACC - Classes poll
Inv_Classes_acp <- Inv_Classes %>%
column_to_rownames(var = "Site_gestion_date")
Y <- Inv_Classes_acp
X <- Inv %>%
select(!c(Site, Parcelle,Gestion_2, Gestion_4, Gestion_moment_4, Date, Heure_debut, Heure_fin, Quartier, Meteo, Jours, Mixte_isole, Activite, Nombre_quadrats))
#dim(X) ; dim(Y)
X<-X[rowSums(Y)!=0,]
Y<-Y[rowSums(Y)!=0,]
Y<-Y[,colSums(Y)!=0]
n <- nrow(Y)
#n == nrow(X)
p <- ncol(Y)
m <- ncol(X)
gestion_2 <- as.factor(X$Gestion_2)
gestion_5 <- as.factor(X$Gestion_moment_5)
afc_Y <- dudi.coa(Y, scannf = FALSE, nf = p)
acc <- pcaiv(afc_Y, X, scannf = FALSE, nf = m)
Q <- viz_coaiv(acc,"co")
viz_coaiv(acc,"li")
viz_coaiv(acc, "li", axes = c(2,3))
Z <- viz_coaiv(acc, "ls")
viz_coaiv(acc, "match")
viz_coaiv(acc, "fa") 
E <- viz_coaiv(acc, "cor")
viz_coaiv(acc, "cor", axes=c(1,3)) 
viz_coaiv(acc, "li", axes = c(1,3))
R <- viz_coaiv(acc, "as")
viz_coaiv(acc, "as", axes=c(2,3)) 
Z = Z + aes(color = gestion_5) + labs(title = "Position des parcelles", color = "Type de gestion", fill = "Type de gestion", shape = "Type de gestion") +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c"))
Q = Q + labs(title = "Position des catégories de pollinisateurs")
E= E + labs(title = "Variables explicatives")
R= R + labs(title = "Participations des axes à la dimention 1-2")
(Q + E) / (Z + R)
3.1.2.4 ACC - Fin juillet
fin_juil_espPl<- FinJuil_AllPl[,c(1,33:82)] %>%
column_to_rownames(var = "Site_gestion_date")
Y <- fin_juil_espPl
X <- Inv %>%
filter(Periode=="Fin juillet") %>%
select(!c(Site, Parcelle, Gestion_2, Gestion_4, Gestion_moment_4, Date, Heure_debut, Heure_fin, Jours, Periode, Quartier, Meteo,Jours, Mixte_isole, Activite, Nombre_quadrats))
# dim(X) ; dim(Y)
X<-X[rowSums(Y)!=0,]
Y<-Y[rowSums(Y)!=0,]
Y<-Y[,colSums(Y)!=0]
n <- nrow(Y)
# n == nrow(X)
p <- ncol(Y)
m <- ncol(X)
gestion_2 <- as.factor(X$Gestion_2)
gestion_5 <- as.factor(X$Gestion_moment_5)
afc_Y <- dudi.coa(Y, scannf = FALSE, nf = p)
acc <- pcaiv(afc_Y, X, scannf = FALSE, nf = m)
Q <- viz_coaiv(acc,"co")
viz_coaiv(acc,"li")
viz_coaiv(acc, "li", axes = c(2,3))
Z <- viz_coaiv(acc, "ls")
viz_coaiv(acc, "match")
viz_coaiv(acc, "fa") 
E <- viz_coaiv(acc, "cor")
viz_coaiv(acc, "cor", axes=c(1,3)) 
viz_coaiv(acc, "li", axes = c(1,3))
R <- viz_coaiv(acc, "as")
viz_coaiv(acc, "as", axes=c(2,3)) 
Z = Z + aes(color = gestion_5) + labs(title = "Position des parcelles", color = "Type de gestion", fill = "Type de gestion", shape = "Type de gestion") +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c"))
Q = Q + labs(title = "Position des espèces de plantes")
E= E + labs(title = "Variables explicatives")
R= R + labs(title = "Participations des axes à la dimention 1-2")
(Q + E) / (Z + R)
3.2 Modèles linéaires
3.2.1 LMM 1. G2 - Sélection et simplification des modèles
3.2.1.1 S Plant
LM1_S_Pl <- lmer(S_Plant ~ Area_gis_m_sq + Temperature + Periode * Gestion_2 + (1|Parcelle), data = Inv)
# check_model(LM1_S_Pl)
# shapiro.test(residuals(LM1_S_Pl))
LM1_S_Pl <- lmer(sqrt(S_Plant) ~ Area_gis_m_sq + Temperature + Periode * Gestion_2 + (1|Parcelle), data = Inv)
# check_model(LM1_S_Pl)
# shapiro.test(residuals(LM1_S_Pl))
step(LM1_S_Pl, direction = "backward") # sqrt(S_Plant) ~ Gestion_2 + (1 | Parcelle)check_model(LM1_S_Pl)
Anova(LM1_S_Pl)## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: sqrt(S_Plant)
## Chisq Df Pr(>Chisq)
## Area_gis_m_sq 0.6372 1 0.424721
## Temperature 2.7291 1 0.098534 .
## Periode 4.9930 2 0.082372 .
## Gestion_2 6.9948 1 0.008175 **
## Periode:Gestion_2 1.4705 2 0.479388
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1summary(LM1_S_Pl)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: sqrt(S_Plant) ~ Area_gis_m_sq + Temperature + Periode * Gestion_2 +
## (1 | Parcelle)
## Data: Inv
##
## REML criterion at convergence: 263.7
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.69122 -0.41288 0.01216 0.50325 2.37157
##
## Random effects:
## Groups Name Variance Std.Dev.
## Parcelle (Intercept) 0.3987 0.6314
## Residual 0.1617 0.4021
## Number of obs: 134, groups: Parcelle, 46
##
## Fixed effects:
## Estimate Std. Error df t value
## (Intercept) 2.995e+00 4.097e-01 1.202e+02 7.309
## Area_gis_m_sq -3.186e-05 3.991e-05 4.201e+01 -0.798
## Temperature -2.617e-02 1.584e-02 1.021e+02 -1.652
## PeriodeMi-juillet -1.233e-01 1.242e-01 8.178e+01 -0.992
## PeriodeFin juillet -2.525e-01 1.347e-01 8.358e+01 -1.875
## Gestion_2Tonte -5.624e-01 2.276e-01 6.078e+01 -2.471
## PeriodeMi-juillet:Gestion_2Tonte -6.997e-02 1.687e-01 8.183e+01 -0.415
## PeriodeFin juillet:Gestion_2Tonte 1.426e-01 1.786e-01 8.353e+01 0.798
## Pr(>|t|)
## (Intercept) 3.25e-11 ***
## Area_gis_m_sq 0.4292
## Temperature 0.1016
## PeriodeMi-juillet 0.3240
## PeriodeFin juillet 0.0643 .
## Gestion_2Tonte 0.0163 *
## PeriodeMi-juillet:Gestion_2Tonte 0.6794
## PeriodeFin juillet:Gestion_2Tonte 0.4270
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) Ar_g__ Tmprtr PrdM-j PrdFnj Gst_2T PM-:G_
## Are_gs_m_sq -0.206
## Temperature -0.910 0.103
## PeridM-jllt -0.107 -0.005 -0.049
## PeridFnjllt -0.231 0.018 0.100 0.455
## Gestin_2Tnt -0.308 -0.216 0.052 0.270 0.255
## PrdM-j:G_2T 0.175 -0.007 -0.069 -0.731 -0.346 -0.372
## PrdFjl:G_2T 0.268 -0.024 -0.178 -0.338 -0.764 -0.356 0.482
## fit warnings:
## Some predictor variables are on very different scales: consider rescaling# LM1_S_Pl
eLM1_S_Pl_P<-emmeans(LM1_S_Pl,"Periode")
mcLM1_S_Pl_P<-cld(eLM1_S_Pl_P,ajust="tukey")
mcLM1_S_Pl_P$.group<-as.numeric(mcLM1_S_Pl_P$.group)
mcLM1_S_Pl_P$group[mcLM1_S_Pl_P$.group == 1] <- "a"
mcLM1_S_Pl_P$group[mcLM1_S_Pl_P$.group == 2] <- "b"
mcLM1_S_Pl_P## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 1.87 0.114 68.5 1.64 2.10 1 a
## Mi-juillet 1.89 0.111 63.5 1.67 2.11 1 a
## Juin 2.05 0.111 63.5 1.83 2.27 1 a
##
## Results are averaged over the levels of: Gestion_2
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.eLM1_S_Pl_G2<-emmeans(LM1_S_Pl,"Gestion_2")
mcLM1_S_Pl_G2<-cld(eLM1_S_Pl_G2,ajust="tukey")
mcLM1_S_Pl_G2$.group<-as.numeric(mcLM1_S_Pl_G2$.group)
mcLM1_S_Pl_G2$group[mcLM1_S_Pl_G2$.group == 1] <- "a"
mcLM1_S_Pl_G2$group[mcLM1_S_Pl_G2$.group == 2] <- "b"
mcLM1_S_Pl_G2## Gestion_2 emmean SE df lower.CL upper.CL .group group
## Tonte 1.67 0.137 42.4 1.39 1.94 1 a
## Fauche 2.21 0.150 43.5 1.90 2.51 2 b
##
## Results are averaged over the levels of: Periode
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.3.2.1.2 Ab Plant
LM1_Ab_Pl <- lmer(Ab_Plant ~ Area_gis_m_sq + Temperature + Periode * Gestion_2 + (1|Parcelle), data = Inv)
# check_model(LM1_Ab_Pl)
# shapiro.test(residuals(LM1_Ab_Pl))
LM1_Ab_Pl <- lmer(sqrt(Ab_Plant) ~ Area_gis_m_sq + Temperature + Periode * Gestion_2 + (1|Parcelle), data = Inv)
# check_model(LM1_Ab_Pl)
# shapiro.test(residuals(LM1_Ab_Pl))
step(LM1_Ab_Pl, direction = "backward") # sqrt(Ab_Plant) ~ Temperature + Periode + (1 | Parcelle)check_model(LM1_Ab_Pl)
Anova(LM1_Ab_Pl)## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: sqrt(Ab_Plant)
## Chisq Df Pr(>Chisq)
## Area_gis_m_sq 0.0043 1 0.94759
## Temperature 4.7783 1 0.02882 *
## Periode 31.3188 2 1.582e-07 ***
## Gestion_2 0.6270 1 0.42847
## Periode:Gestion_2 3.1873 2 0.20319
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# summary(LM1_Ab_Pl)
coef(LM1_Ab_Pl)## (Intercept) Area_gis_m_sq
## 1.508921e+01 1.430211e-05
## Temperature PeriodeMi-juillet
## -2.382167e-01 -1.492381e+00
## PeriodeFin juillet Gestion_2Tonte
## -2.734592e+00 1.369802e-01
## PeriodeMi-juillet:Gestion_2Tonte PeriodeFin juillet:Gestion_2Tonte
## -2.149672e+00 -9.127128e-01# LM1_Ab_Pl
eLM1_Ab_Pl_P<-emmeans(LM1_Ab_Pl,"Periode")
mcLM1_Ab_Pl_P<-cld(eLM1_Ab_Pl_P,ajust="tukey")
mcLM1_Ab_Pl_P$.group<-as.numeric(mcLM1_Ab_Pl_P$.group)
mcLM1_Ab_Pl_P$group[mcLM1_Ab_Pl_P$.group == 1] <- "b"
mcLM1_Ab_Pl_P$group[mcLM1_Ab_Pl_P$.group == 2] <- "a"
mcLM1_Ab_Pl_P## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 6.50 0.668 84.8 5.17 7.83 1 b
## Mi-juillet 7.12 0.645 77.9 5.84 8.41 1 b
## Juin 9.69 0.645 77.9 8.41 10.97 2 a
##
## Results are averaged over the levels of: Gestion_2
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.eLM1_Ab_Pl_G2<-emmeans(LM1_Ab_Pl,"Gestion_2")
mcLM1_Ab_Pl_G2<-cld(eLM1_Ab_Pl_G2,ajust="tukey")
mcLM1_Ab_Pl_G2$.group<-as.numeric(mcLM1_Ab_Pl_G2$.group)
mcLM1_Ab_Pl_G2$group[mcLM1_Ab_Pl_G2$.group == 1] <- "a"
mcLM1_Ab_Pl_G2$group[mcLM1_Ab_Pl_G2$.group == 2] <- "b"
mcLM1_Ab_Pl_G2## Gestion_2 emmean SE df lower.CL upper.CL .group group
## Tonte 7.33 0.742 42.0 5.83 8.82 1 a
## Fauche 8.21 0.821 43.7 6.56 9.87 1 a
##
## Results are averaged over the levels of: Periode
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.3.2.1.3 S Poll
LM1_S_Po <- lmer(S_Poll ~ Area_gis_m_sq + Temperature + Periode * Gestion_2 + (1|Parcelle), data = Inv)
# check_model(LM1_S_Po)
# shapiro.test(residuals(LM1_S_Po))
LM1_S_Po <- lmer(sqrt(S_Poll) ~ Area_gis_m_sq + Temperature + Periode * Gestion_2 + (1|Parcelle), data = Inv)
# check_model(LM1_S_Po)
# shapiro.test(residuals(LM1_S_Po))
step(LM1_S_Po, direction = "backward") # sqrt(S_Poll) ~ Gestion_2 + (1 | Parcelle)check_model(LM1_S_Po)
Anova(LM1_S_Po)## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: sqrt(S_Poll)
## Chisq Df Pr(>Chisq)
## Area_gis_m_sq 0.3284 1 0.56659
## Temperature 0.7725 1 0.37945
## Periode 5.1152 2 0.07749 .
## Gestion_2 32.0557 1 1.498e-08 ***
## Periode:Gestion_2 0.1999 2 0.90489
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# summary(LM1_S_Po)
# LM1_S_Po
eLM1_S_Po_P<-emmeans(LM1_S_Po,"Periode")
mcLM1_S_Po_P<-cld(eLM1_S_Po_P,ajust="tukey")
mcLM1_S_Po_P$.group<-as.numeric(mcLM1_S_Po_P$.group)
mcLM1_S_Po_P$group[mcLM1_S_Po_P$.group == 1] <- "a"
mcLM1_S_Po_P$group[mcLM1_S_Po_P$.group == 2] <- "b"
mcLM1_S_Po_P## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 2.01 0.142 97.7 1.72 2.29 1 a
## Mi-juillet 2.04 0.136 90.6 1.78 2.31 1 a
## Juin 2.30 0.136 90.6 2.03 2.57 1 a
##
## Results are averaged over the levels of: Gestion_2
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.eLM1_S_Po_G2<-emmeans(LM1_S_Po,"Gestion_2")
mcLM1_S_Po_G2<-cld(eLM1_S_Po_G2,ajust="tukey")
mcLM1_S_Po_G2$.group<-as.numeric(mcLM1_S_Po_G2$.group)
mcLM1_S_Po_G2$group[mcLM1_S_Po_G2$.group == 1] <- "a"
mcLM1_S_Po_G2$group[mcLM1_S_Po_G2$.group == 2] <- "b"
mcLM1_S_Po_G2## Gestion_2 emmean SE df lower.CL upper.CL .group group
## Tonte 1.48 0.147 41.6 1.19 1.78 1 a
## Fauche 2.75 0.164 43.9 2.42 3.08 2 b
##
## Results are averaged over the levels of: Periode
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.3.2.1.4 Ab Poll
LM1_Ab_Po <- lmer(Ab_Poll ~ Area_gis_m_sq + Temperature + Periode * Gestion_2 + (1|Parcelle), data = Inv)
# check_model(LM1_Ab_Po)
# shapiro.test(residuals(LM1_Ab_Po))
LM1_Ab_Po <- lmer(sqrt(Ab_Poll) ~ Area_gis_m_sq + Temperature + Periode * Gestion_2 + (1|Parcelle), data = Inv)
# check_model(LM1_Ab_Po)
# shapiro.test(residuals(LM1_Ab_Po))
step(LM1_Ab_Po, direction = "backward") # sqrt(Ab_Poll) ~ Gestion_2 + (1 | Parcelle)check_model(LM1_Ab_Po)
Anova(LM1_Ab_Po)## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: sqrt(Ab_Poll)
## Chisq Df Pr(>Chisq)
## Area_gis_m_sq 0.0360 1 0.8496
## Temperature 0.0007 1 0.9791
## Periode 3.3492 2 0.1874
## Gestion_2 28.7659 1 8.168e-08 ***
## Periode:Gestion_2 2.9264 2 0.2315
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# summary(LM1_Ab_Po)
# LM1_Ab_Po
eLM1_Ab_Po_P<-emmeans(LM1_Ab_Po,"Periode")
mcLM1_Ab_Po_P<-cld(eLM1_Ab_Po_P,ajust="tukey")
mcLM1_Ab_Po_P$.group<-as.numeric(mcLM1_Ab_Po_P$.group)
mcLM1_Ab_Po_P$group[mcLM1_Ab_Po_P$.group == 1] <- "a"
mcLM1_Ab_Po_P$group[mcLM1_Ab_Po_P$.group == 2] <- "b"
mcLM1_Ab_Po_P## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 2.81 0.254 100.8 2.30 3.31 1 a
## Mi-juillet 3.18 0.243 93.8 2.69 3.66 1 a
## Juin 3.30 0.243 93.9 2.82 3.78 1 a
##
## Results are averaged over the levels of: Gestion_2
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.eLM1_Ab_Po_G2<-emmeans(LM1_Ab_Po,"Gestion_2")
mcLM1_Ab_Po_G2<-cld(eLM1_Ab_Po_G2,ajust="tukey")
mcLM1_Ab_Po_G2$.group<-as.numeric(mcLM1_Ab_Po_G2$.group)
mcLM1_Ab_Po_G2$group[mcLM1_Ab_Po_G2$.group == 1] <- "a"
mcLM1_Ab_Po_G2$group[mcLM1_Ab_Po_G2$.group == 2] <- "b"
mcLM1_Ab_Po_G2## Gestion_2 emmean SE df lower.CL upper.CL .group group
## Tonte 2.04 0.259 41.5 1.52 2.57 1 a
## Fauche 4.14 0.289 44.0 3.56 4.73 2 b
##
## Results are averaged over the levels of: Periode
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.3.2.2 LMM 2. G5 - Sélection et simplification des modèles
3.2.2.1 S Plant
LM2_S_Pl <- lmer(S_Plant ~ Area_gis_m_sq + Temperature + Periode * Gestion_moment_5 + (1|Parcelle), data = Inv)
# check_model(LM2_S_Pl)
# shapiro.test(residuals(LM2_S_Pl))
LM2_S_Pl <- lmer(sqrt(S_Plant) ~ Area_gis_m_sq + Temperature + Periode * Gestion_moment_5 + (1|Parcelle), data = Inv)
# check_model(LM2_S_Pl)
# shapiro.test(residuals(LM2_S_Pl))
step(LM2_S_Pl, direction = "backward") # sqrt(S_Plant) ~ Periode + Gestion_moment_5 + (1 | Parcelle)check_model(LM2_S_Pl)
Anova(LM2_S_Pl)## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: sqrt(S_Plant)
## Chisq Df Pr(>Chisq)
## Area_gis_m_sq 0.1796 1 0.6717
## Temperature 1.1648 1 0.2805
## Periode 7.6610 2 0.0217 *
## Gestion_moment_5 74.9761 4 2.016e-15 ***
## Periode:Gestion_moment_5 11.2378 8 0.1886
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# summary(LM2_S_Pl)
# LM2_S_Pl
eLM2_S_Pl_P<-emmeans(LM2_S_Pl,"Periode")
mcLM2_S_Pl_P<-cld(eLM2_S_Pl_P,ajust="tukey")
mcLM2_S_Pl_P$.group<-as.numeric(mcLM2_S_Pl_P$.group)
mcLM2_S_Pl_P$group[mcLM2_S_Pl_P$.group == 1] <- "a"
mcLM2_S_Pl_P$group[mcLM2_S_Pl_P$.group == 2] <- "b"
mcLM2_S_Pl_P## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 1.84 0.1008 82.7 1.64 2.04 1 a
## Mi-juillet 1.85 0.0930 67.9 1.67 2.04 1 a
## Juin 1.98 0.0939 69.7 1.79 2.17 1 a
##
## Results are averaged over the levels of: Gestion_moment_5
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Pl_P_G <- Inv %>%
filter(Gestion_moment_5 == "Graminées")
LM2_S_Pl_P_G <- lmer(sqrt(S_Plant) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_S_Pl_P_G)
eLM2_S_Pl_P_G<-emmeans(LM2_S_Pl_P_G,"Periode")
mcLM2_S_Pl_P_G<-cld(eLM2_S_Pl_P_G,ajust="tukey")
mcLM2_S_Pl_P_G$.group<-as.numeric(mcLM2_S_Pl_P_G$.group)
mcLM2_S_Pl_P_G$group[mcLM2_S_Pl_P_G$.group == 1] <- "a"
mcLM2_S_Pl_P_G$group[mcLM2_S_Pl_P_G$.group == 2] <- "b"
mcLM2_S_Pl_P_G## Periode emmean SE df lower.CL upper.CL .group group
## Mi-juillet 1.05 0.384 2.71 -0.254 2.35 1 a
## Juin 1.07 0.387 2.76 -0.221 2.37 1 a
## Fin juillet 1.37 0.399 3.08 0.119 2.62 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Pl_P_F <- Inv %>%
filter(Gestion_moment_5 == "Fleuri")
LM2_S_Pl_P_F <- lmer(sqrt(S_Plant) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_S_Pl_P_F)
eLM2_S_Pl_P_F<-emmeans(LM2_S_Pl_P_F,"Periode")
mcLM2_S_Pl_P_F<-cld(eLM2_S_Pl_P_F,ajust="tukey")
mcLM2_S_Pl_P_F$.group<-as.numeric(mcLM2_S_Pl_P_F$.group)
mcLM2_S_Pl_P_F$group[mcLM2_S_Pl_P_F$.group == 1] <- "a"
mcLM2_S_Pl_P_F$group[mcLM2_S_Pl_P_F$.group == 2] <- "b"
mcLM2_S_Pl_P_F## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 1.93 0.186 14.5 1.53 2.33 1 a
## Mi-juillet 2.31 0.175 12.1 1.92 2.69 2 b
## Juin 2.47 0.175 12.1 2.08 2.85 2 b
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Pl_P_S <- Inv %>%
filter(Gestion_moment_5 == "Semé")
LM2_S_Pl_P_S <- lmer(sqrt(S_Plant) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_S_Pl_P_S)
eLM2_S_Pl_P_S<-emmeans(LM2_S_Pl_P_S,"Periode")
mcLM2_S_Pl_P_S<-cld(eLM2_S_Pl_P_S,ajust="tukey")
mcLM2_S_Pl_P_S$.group<-as.numeric(mcLM2_S_Pl_P_S$.group)
mcLM2_S_Pl_P_S$group[mcLM2_S_Pl_P_S$.group == 1] <- "a"
mcLM2_S_Pl_P_S$group[mcLM2_S_Pl_P_S$.group == 2] <- "b"
mcLM2_S_Pl_P_S## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 2.80 0.183 8.58 2.39 3.22 1 a
## Mi-juillet 2.97 0.178 8.19 2.56 3.38 1 a
## Juin 3.06 0.177 8.12 2.65 3.47 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Pl_P_Tr <- Inv %>%
filter(Gestion_moment_5 == "Tonte récente")
LM2_S_Pl_P_Tr <- lmer(sqrt(S_Plant) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_S_Pl_P_Tr)
eLM2_S_Pl_P_Tr<-emmeans(LM2_S_Pl_P_Tr,"Periode")
mcLM2_S_Pl_P_Tr<-cld(eLM2_S_Pl_P_Tr,ajust="tukey")
mcLM2_S_Pl_P_Tr$.group<-as.numeric(mcLM2_S_Pl_P_Tr$.group)
mcLM2_S_Pl_P_Tr$group[mcLM2_S_Pl_P_Tr$.group == 1] <- "a"
mcLM2_S_Pl_P_Tr$group[mcLM2_S_Pl_P_Tr$.group == 2] <- "b"
mcLM2_S_Pl_P_Tr## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 1.18 0.245 19.8 0.664 1.69 1 a
## Mi-juillet 1.22 0.136 21.0 0.939 1.50 1 a
## Juin 1.54 0.182 20.9 1.158 1.92 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Pl_P_Tt <- Inv %>%
filter(Gestion_moment_5 == "Tonte tardive")
LM2_S_Pl_P_Tt <- lmer(sqrt(S_Plant) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_S_Pl_P_Tt)
eLM2_S_Pl_P_Tt<-emmeans(LM2_S_Pl_P_Tt,"Periode")
mcLM2_S_Pl_P_Tt<-cld(eLM2_S_Pl_P_Tt,ajust="tukey")
mcLM2_S_Pl_P_Tt$.group<-as.numeric(mcLM2_S_Pl_P_Tt$.group)
mcLM2_S_Pl_P_Tt$group[mcLM2_S_Pl_P_Tt$.group == 1] <- "a"
mcLM2_S_Pl_P_Tt$group[mcLM2_S_Pl_P_Tt$.group == 2] <- "b"
mcLM2_S_Pl_P_Tt## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 1.72 0.0882 43.2 1.54 1.90 1 a
## Juin 1.90 0.0992 43.4 1.70 2.10 1 a
## Mi-juillet 1.91 0.1122 44.0 1.69 2.14 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.eLM2_S_Pl_G5<-emmeans(LM2_S_Pl,"Gestion_moment_5")
mcLM2_S_Pl_G5<-cld(eLM2_S_Pl_G5,ajust="tukey")
mcLM2_S_Pl_G5$.group<-as.numeric(mcLM2_S_Pl_G5$.group)
mcLM2_S_Pl_G5$group[mcLM2_S_Pl_G5$.group == 1] <- "a"
mcLM2_S_Pl_G5$group[mcLM2_S_Pl_G5$.group == 2] <- "b"
mcLM2_S_Pl_G5$group[mcLM2_S_Pl_G5$.group == 3] <- "c"
mcLM2_S_Pl_G5$group[mcLM2_S_Pl_G5$.group == 23] <- "bc"
mcLM2_S_Pl_G5## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Graminées 1.06 0.249 41.8 0.563 1.57 1 a
## Tonte récente 1.37 0.123 77.4 1.127 1.62 1 a
## Tonte tardive 1.84 0.106 49.6 1.627 2.05 2 b
## Fleuri 2.24 0.152 42.0 1.937 2.55 23 bc
## Semé 2.93 0.203 39.7 2.516 3.34 3 c
##
## Results are averaged over the levels of: Periode
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.3.2.2.2 Ab Plant
LM2_Ab_Pl <- lmer(Ab_Plant ~ Area_gis_m_sq + Temperature + Periode * Gestion_moment_5 + (1|Parcelle), data = Inv)
# check_model(LM2_Ab_Pl)
# shapiro.test(residuals(LM2_Ab_Pl))
LM2_Ab_Pl <- lmer(sqrt(Ab_Plant) ~ Area_gis_m_sq + Temperature + Periode * Gestion_moment_5 + (1|Parcelle), data = Inv)
# check_model(LM2_Ab_Pl)
# shapiro.test(residuals(LM2_Ab_Pl))
step(LM2_Ab_Pl, direction = "backward") # sqrt(Ab_Plant) ~ Temperature + Periode + Gestion_moment_5 + (1 | Parcelle)check_model(LM2_Ab_Pl)
Anova(LM2_Ab_Pl)## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: sqrt(Ab_Plant)
## Chisq Df Pr(>Chisq)
## Area_gis_m_sq 0.2566 1 0.6125
## Temperature 1.8593 1 0.1727
## Periode 41.7713 2 8.501e-10 ***
## Gestion_moment_5 44.9290 4 4.113e-09 ***
## Periode:Gestion_moment_5 7.1898 8 0.5163
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# summary(LM2_Ab_Pl)
# LM2_Ab_Pl
eLM2_Ab_Pl_P<-emmeans(LM2_Ab_Pl,"Periode")
mcLM2_Ab_Pl_P<-cld(eLM2_Ab_Pl_P,ajust="tukey")
mcLM2_Ab_Pl_P$.group<-as.numeric(mcLM2_Ab_Pl_P$.group)
mcLM2_Ab_Pl_P$group[mcLM2_Ab_Pl_P$.group == 1] <- "b"
mcLM2_Ab_Pl_P$group[mcLM2_Ab_Pl_P$.group == 2] <- "a"
mcLM2_Ab_Pl_P## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 6.07 0.670 92.8 4.74 7.40 1 b
## Mi-juillet 6.94 0.609 77.0 5.72 8.15 1 b
## Juin 8.78 0.616 79.0 7.55 10.00 2 a
##
## Results are averaged over the levels of: Gestion_moment_5
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_Ab_Pl_P_G <- Inv %>%
filter(Gestion_moment_5 == "Graminées")
LM2_Ab_Pl_P_G <- lmer(sqrt(S_Plant) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_Ab_Pl_P_G)
eLM2_Ab_Pl_P_G<-emmeans(LM2_Ab_Pl_P_G,"Periode")
mcLM2_Ab_Pl_P_G<-cld(eLM2_Ab_Pl_P_G,ajust="tukey")
mcLM2_Ab_Pl_P_G$.group<-as.numeric(mcLM2_Ab_Pl_P_G$.group)
mcLM2_Ab_Pl_P_G$group[mcLM2_Ab_Pl_P_G$.group == 1] <- "a"
mcLM2_Ab_Pl_P_G$group[mcLM2_Ab_Pl_P_G$.group == 2] <- "b"
mcLM2_Ab_Pl_P_G## Periode emmean SE df lower.CL upper.CL .group group
## Mi-juillet 1.05 0.384 2.71 -0.254 2.35 1 a
## Juin 1.07 0.387 2.76 -0.221 2.37 1 a
## Fin juillet 1.37 0.399 3.08 0.119 2.62 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_Ab_Pl_P_F <- Inv %>%
filter(Gestion_moment_5 == "Fleuri")
LM2_Ab_Pl_P_F <- lmer(sqrt(S_Plant) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_Ab_Pl_P_F)
eLM2_Ab_Pl_P_F<-emmeans(LM2_Ab_Pl_P_F,"Periode")
mcLM2_Ab_Pl_P_F<-cld(eLM2_Ab_Pl_P_F,ajust="tukey")
mcLM2_Ab_Pl_P_F$.group<-as.numeric(mcLM2_Ab_Pl_P_F$.group)
mcLM2_Ab_Pl_P_F$group[mcLM2_Ab_Pl_P_F$.group == 1] <- "a"
mcLM2_Ab_Pl_P_F$group[mcLM2_Ab_Pl_P_F$.group == 2] <- "b"
mcLM2_Ab_Pl_P_F## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 1.93 0.186 14.5 1.53 2.33 1 a
## Mi-juillet 2.31 0.175 12.1 1.92 2.69 2 b
## Juin 2.47 0.175 12.1 2.08 2.85 2 b
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_Ab_Pl_P_S <- Inv %>%
filter(Gestion_moment_5 == "Semé")
LM2_Ab_Pl_P_S <- lmer(sqrt(S_Plant) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_Ab_Pl_P_S)
eLM2_Ab_Pl_P_S<-emmeans(LM2_Ab_Pl_P_S,"Periode")
mcLM2_Ab_Pl_P_S<-cld(eLM2_Ab_Pl_P_S,ajust="tukey")
mcLM2_Ab_Pl_P_S$.group<-as.numeric(mcLM2_Ab_Pl_P_S$.group)
mcLM2_Ab_Pl_P_S$group[mcLM2_Ab_Pl_P_S$.group == 1] <- "a"
mcLM2_Ab_Pl_P_S$group[mcLM2_Ab_Pl_P_S$.group == 2] <- "b"
mcLM2_Ab_Pl_P_S## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 2.80 0.183 8.58 2.39 3.22 1 a
## Mi-juillet 2.97 0.178 8.19 2.56 3.38 1 a
## Juin 3.06 0.177 8.12 2.65 3.47 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_Ab_Pl_P_Tr <- Inv %>%
filter(Gestion_moment_5 == "Tonte récente")
LM2_Ab_Pl_P_Tr <- lmer(sqrt(S_Plant) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_Ab_Pl_P_Tr)
eLM2_Ab_Pl_P_Tr<-emmeans(LM2_Ab_Pl_P_Tr,"Periode")
mcLM2_Ab_Pl_P_Tr<-cld(eLM2_Ab_Pl_P_Tr,ajust="tukey")
mcLM2_Ab_Pl_P_Tr$.group<-as.numeric(mcLM2_Ab_Pl_P_Tr$.group)
mcLM2_Ab_Pl_P_Tr$group[mcLM2_Ab_Pl_P_Tr$.group == 1] <- "a"
mcLM2_Ab_Pl_P_Tr$group[mcLM2_Ab_Pl_P_Tr$.group == 2] <- "b"
mcLM2_Ab_Pl_P_Tr## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 1.18 0.245 19.8 0.664 1.69 1 a
## Mi-juillet 1.22 0.136 21.0 0.939 1.50 1 a
## Juin 1.54 0.182 20.9 1.158 1.92 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_Ab_Pl_P_Tt <- Inv %>%
filter(Gestion_moment_5 == "Tonte tardive")
LM2_Ab_Pl_P_Tt <- lmer(sqrt(S_Plant) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_Ab_Pl_P_Tt)
eLM2_Ab_Pl_P_Tt<-emmeans(LM2_Ab_Pl_P_Tt,"Periode")
mcLM2_Ab_Pl_P_Tt<-cld(eLM2_Ab_Pl_P_Tt,ajust="tukey")
mcLM2_Ab_Pl_P_Tt$.group<-as.numeric(mcLM2_Ab_Pl_P_Tt$.group)
mcLM2_Ab_Pl_P_Tt$group[mcLM2_Ab_Pl_P_Tt$.group == 1] <- "a"
mcLM2_Ab_Pl_P_Tt$group[mcLM2_Ab_Pl_P_Tt$.group == 2] <- "b"
mcLM2_Ab_Pl_P_Tt## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 1.72 0.0882 43.2 1.54 1.90 1 a
## Juin 1.90 0.0992 43.4 1.70 2.10 1 a
## Mi-juillet 1.91 0.1122 44.0 1.69 2.14 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.eLM2_Ab_Pl_G5<-emmeans(LM2_Ab_Pl,"Gestion_moment_5")
mcLM2_Ab_Pl_G5<-cld(eLM2_Ab_Pl_G5,ajust="tukey")
mcLM2_Ab_Pl_G5$.group<-as.numeric(mcLM2_Ab_Pl_G5$.group)
mcLM2_Ab_Pl_G5$group[mcLM2_Ab_Pl_G5$.group == 1] <- "a"
mcLM2_Ab_Pl_G5$group[mcLM2_Ab_Pl_G5$.group == 2] <- "b"
mcLM2_Ab_Pl_G5$group[mcLM2_Ab_Pl_G5$.group == 3] <- "c"
mcLM2_Ab_Pl_G5$group[mcLM2_Ab_Pl_G5$.group == 23] <- "bc"
mcLM2_Ab_Pl_G5## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Graminées 2.98 1.564 42.0 -0.174 6.14 1 a
## Tonte récente 5.11 0.816 85.1 3.483 6.73 1 a
## Tonte tardive 8.60 0.675 52.3 7.246 9.96 2 b
## Fleuri 8.72 0.956 42.4 6.793 10.65 2 b
## Semé 10.89 1.269 39.4 8.323 13.45 2 b
##
## Results are averaged over the levels of: Periode
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.3.2.2.3 S Poll
LM2_S_Po <- lmer(S_Poll ~ Area_gis_m_sq + Temperature + Periode * Gestion_moment_5 + (1|Parcelle), data = Inv)
# check_model(LM2_S_Po)
# shapiro.test(residuals(LM2_S_Po))
LM2_S_Po <- lmer(sqrt(S_Poll) ~ Area_gis_m_sq + Temperature + Periode * Gestion_moment_5 + (1|Parcelle), data = Inv)
# check_model(LM2_S_Po)
# shapiro.test(residuals(LM2_S_Po))
step(LM2_S_Po, direction = "backward")
# sqrt(S_Poll) ~ Periode + Gestion_moment_5 + (1 | Parcelle) + Periode:Gestion_moment_5check_model(LM2_S_Po)
Anova(LM2_S_Po)## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: sqrt(S_Poll)
## Chisq Df Pr(>Chisq)
## Area_gis_m_sq 0.0491 1 0.82467
## Temperature 0.3952 1 0.52957
## Periode 7.6092 2 0.02227 *
## Gestion_moment_5 93.2565 4 < 2e-16 ***
## Periode:Gestion_moment_5 18.1967 8 0.01980 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# summary(LM2_S_Po)
# LM2_S_Po
eLM2_S_Po_P<-emmeans(LM2_S_Po,"Periode")
mcLM2_S_Po_P<-cld(eLM2_S_Po_P,ajust="tukey")
mcLM2_S_Po_P$.group<-as.numeric(mcLM2_S_Po_P$.group)
mcLM2_S_Po_P$group[mcLM2_S_Po_P$.group == 1] <- "a"
mcLM2_S_Po_P$group[mcLM2_S_Po_P$.group == 2] <- "b"
mcLM2_S_Po_P## Periode emmean SE df lower.CL upper.CL .group group
## Mi-juillet 2.08 0.123 94.0 1.84 2.33 1 a
## Fin juillet 2.11 0.139 106.5 1.83 2.38 1 a
## Juin 2.24 0.125 95.8 1.99 2.48 1 a
##
## Results are averaged over the levels of: Gestion_moment_5
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Po_P_G <- Inv %>%
filter(Gestion_moment_5 == "Graminées")
LM2_S_Po_P_G <- lmer(sqrt(S_Poll) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_S_Po_P_G)
eLM2_S_Po_P_G<-emmeans(LM2_S_Po_P_G,"Periode")
mcLM2_S_Po_P_G<-cld(eLM2_S_Po_P_G,ajust="tukey")
mcLM2_S_Po_P_G$.group<-as.numeric(mcLM2_S_Po_P_G$.group)
mcLM2_S_Po_P_G$group[mcLM2_S_Po_P_G$.group == 1] <- "a"
mcLM2_S_Po_P_G$group[mcLM2_S_Po_P_G$.group == 2] <- "b"
mcLM2_S_Po_P_G## Periode emmean SE df lower.CL upper.CL .group group
## Juin 1.15 0.638 5.85 -0.4187 2.72 1 a
## Mi-juillet 1.58 0.619 5.87 0.0596 3.10 1 a
## Fin juillet 2.46 0.711 5.99 0.7142 4.20 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Po_P_F <- Inv %>%
filter(Gestion_moment_5 == "Fleuri")
LM2_S_Po_P_F <- lmer(sqrt(S_Poll) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_S_Po_P_F)
eLM2_S_Po_P_F<-emmeans(LM2_S_Po_P_F,"Periode")
mcLM2_S_Po_P_F<-cld(eLM2_S_Po_P_F,ajust="tukey")
mcLM2_S_Po_P_F$.group<-as.numeric(mcLM2_S_Po_P_F$.group)
mcLM2_S_Po_P_F$group[mcLM2_S_Po_P_F$.group == 1] <- "a"
mcLM2_S_Po_P_F$group[mcLM2_S_Po_P_F$.group == 2] <- "b"
mcLM2_S_Po_P_F## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 2.60 0.255 20.0 2.07 3.13 1 a
## Mi-juillet 2.81 0.227 16.4 2.33 3.29 1 a
## Juin 2.98 0.228 16.5 2.50 3.47 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Po_P_S <- Inv %>%
filter(Gestion_moment_5 == "Semé")
LM2_S_Po_P_S <- lmer(sqrt(S_Poll) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_S_Po_P_S)
eLM2_S_Po_P_S<-emmeans(LM2_S_Po_P_S,"Periode")
mcLM2_S_Po_P_S<-cld(eLM2_S_Po_P_S,ajust="tukey")
mcLM2_S_Po_P_S$.group<-as.numeric(mcLM2_S_Po_P_S$.group)
mcLM2_S_Po_P_S$group[mcLM2_S_Po_P_S$.group == 1] <- "a"
mcLM2_S_Po_P_S$group[mcLM2_S_Po_P_S$.group == 2] <- "b"
mcLM2_S_Po_P_S$group[mcLM2_S_Po_P_S$.group == 12] <- "ab"
mcLM2_S_Po_P_S## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 2.97 0.208 11.3 2.51 3.43 1 a
## Mi-juillet 3.52 0.201 11.0 3.07 3.96 12 ab
## Juin 3.78 0.199 11.0 3.34 4.21 2 b
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Po_P_Tr <- Inv %>%
filter(Gestion_moment_5 == "Tonte récente")
LM2_S_Po_P_Tr <- lmer(sqrt(S_Poll) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_S_Po_P_Tr)
eLM2_S_Po_P_Tr<-emmeans(LM2_S_Po_P_Tr,"Periode")
mcLM2_S_Po_P_Tr<-cld(eLM2_S_Po_P_Tr,ajust="tukey")
mcLM2_S_Po_P_Tr$.group<-as.numeric(mcLM2_S_Po_P_Tr$.group)
mcLM2_S_Po_P_Tr$group[mcLM2_S_Po_P_Tr$.group == 1] <- "a"
mcLM2_S_Po_P_Tr$group[mcLM2_S_Po_P_Tr$.group == 2] <- "b"
mcLM2_S_Po_P_Tr## Periode emmean SE df lower.CL upper.CL .group group
## Mi-juillet 0.874 0.182 21.0 0.495 1.25 1 a
## Fin juillet 1.060 0.329 19.8 0.373 1.75 1 a
## Juin 1.407 0.244 20.9 0.898 1.91 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Po_P_Tt <- Inv %>%
filter(Gestion_moment_5 == "Tonte tardive")
LM2_S_Po_P_Tt <- lmer(sqrt(S_Poll) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_S_Po_P_Tt)
eLM2_S_Po_P_Tt<-emmeans(LM2_S_Po_P_Tt,"Periode")
mcLM2_S_Po_P_Tt<-cld(eLM2_S_Po_P_Tt,ajust="tukey")
mcLM2_S_Po_P_Tt$.group<-as.numeric(mcLM2_S_Po_P_Tt$.group)
mcLM2_S_Po_P_Tt$group[mcLM2_S_Po_P_Tt$.group == 1] <- "a"
mcLM2_S_Po_P_Tt$group[mcLM2_S_Po_P_Tt$.group == 2] <- "b"
mcLM2_S_Po_P_Tt$group[mcLM2_S_Po_P_Tt$.group == 12] <- "ab"
mcLM2_S_Po_P_Tt## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 1.37 0.143 44.0 1.08 1.66 1 a
## Mi-juillet 1.87 0.185 44.0 1.50 2.24 12 ab
## Juin 1.92 0.161 43.7 1.60 2.25 2 b
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.eLM2_S_Po_G5<-emmeans(LM2_S_Po,"Gestion_moment_5")
mcLM2_S_Po_G5<-cld(eLM2_S_Po_G5,ajust="tukey")
mcLM2_S_Po_G5$.group<-as.numeric(mcLM2_S_Po_G5$.group)
mcLM2_S_Po_G5$group[mcLM2_S_Po_G5$.group == 1] <- "a"
mcLM2_S_Po_G5$group[mcLM2_S_Po_G5$.group == 2] <- "b"
mcLM2_S_Po_G5$group[mcLM2_S_Po_G5$.group == 3] <- "c"
mcLM2_S_Po_G5$group[mcLM2_S_Po_G5$.group == 12] <- "ab"
mcLM2_S_Po_G5## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Tonte récente 1.09 0.167 95.1 0.759 1.42 1 a
## Graminées 1.67 0.292 42.3 1.079 2.26 12 ab
## Tonte tardive 1.72 0.130 57.4 1.455 1.98 2 b
## Fleuri 2.83 0.179 43.0 2.466 3.19 3 c
## Semé 3.41 0.235 38.6 2.932 3.88 3 c
##
## Results are averaged over the levels of: Periode
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Po_G5_Ju <- Inv %>%
filter(Periode == "Juin")
LM2_S_Po_G5_Ju <- lm(sqrt(S_Poll) ~ Area_gis_m_sq + Temperature + Gestion_moment_5, data = CM2_S_Po_G5_Ju)
eLM2_S_Po_G5_Ju<-emmeans(LM2_S_Po_G5_Ju,"Gestion_moment_5")
mcLM2_S_Po_G5_Ju<-cld(eLM2_S_Po_G5_Ju,ajust="tukey")
mcLM2_S_Po_G5_Ju$.group<-as.numeric(mcLM2_S_Po_G5_Ju$.group)
mcLM2_S_Po_G5_Ju$group[mcLM2_S_Po_G5_Ju$.group == 1] <- "a"
mcLM2_S_Po_G5_Ju$group[mcLM2_S_Po_G5_Ju$.group == 2] <- "b"
mcLM2_S_Po_G5_Ju## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Graminées 1.13 0.317 39 0.488 1.77 1 a
## Tonte récente 1.33 0.232 39 0.860 1.80 1 a
## Tonte tardive 1.92 0.163 39 1.593 2.25 1 a
## Fleuri 3.06 0.194 39 2.664 3.45 2 b
## Semé 3.75 0.262 39 3.215 4.28 2 b
##
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Po_G5_mJ <- Inv %>%
filter(Periode == "Mi-juillet")
LM2_S_Po_G5_mJ <- lm(sqrt(S_Poll) ~ Area_gis_m_sq + Temperature + Gestion_moment_5, data = CM2_S_Po_G5_mJ)
eLM2_S_Po_G5_mJ<-emmeans(LM2_S_Po_G5_mJ,"Gestion_moment_5")
mcLM2_S_Po_G5_mJ<-cld(eLM2_S_Po_G5_mJ,ajust="tukey")
mcLM2_S_Po_G5_mJ$.group<-as.numeric(mcLM2_S_Po_G5_mJ$.group)
mcLM2_S_Po_G5_mJ$group[mcLM2_S_Po_G5_mJ$.group == 1] <- "a"
mcLM2_S_Po_G5_mJ$group[mcLM2_S_Po_G5_mJ$.group == 12] <- "ab"
mcLM2_S_Po_G5_mJ$group[mcLM2_S_Po_G5_mJ$.group == 23] <- "bc"
mcLM2_S_Po_G5_mJ$group[mcLM2_S_Po_G5_mJ$.group == 34] <- "cd"
mcLM2_S_Po_G5_mJ$group[mcLM2_S_Po_G5_mJ$.group == 4] <- "d"
mcLM2_S_Po_G5_mJ## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Tonte récente 0.888 0.228 39 0.426 1.35 1 a
## Graminées 1.149 0.422 39 0.294 2.00 12 ab
## Tonte tardive 1.948 0.245 39 1.453 2.44 23 bc
## Fleuri 2.807 0.255 39 2.292 3.32 34 cd
## Semé 3.449 0.341 39 2.760 4.14 4 d
##
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Po_G5_fJ <- Inv %>%
filter(Periode == "Fin juillet")
LM2_S_Po_G5_fJ <- lm(sqrt(S_Poll) ~ Area_gis_m_sq + Temperature + Gestion_moment_5, data = CM2_S_Po_G5_fJ)
eLM2_S_Po_G5_fJ<-emmeans(LM2_S_Po_G5_fJ,"Gestion_moment_5")
mcLM2_S_Po_G5_fJ<-cld(eLM2_S_Po_G5_fJ,ajust="tukey")
mcLM2_S_Po_G5_fJ$.group<-as.numeric(mcLM2_S_Po_G5_fJ$.group)
mcLM2_S_Po_G5_fJ$group[mcLM2_S_Po_G5_fJ$.group == 1] <- "a"
mcLM2_S_Po_G5_fJ$group[mcLM2_S_Po_G5_fJ$.group == 2] <- "b"
mcLM2_S_Po_G5_fJ$group[mcLM2_S_Po_G5_fJ$.group == 12] <- "ab"
mcLM2_S_Po_G5_fJ## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Tonte récente 0.99 0.333 35 0.314 1.67 1 a
## Tonte tardive 1.37 0.164 35 1.042 1.71 1 a
## Fleuri 2.61 0.261 35 2.077 3.14 2 b
## Graminées 2.81 0.420 35 1.957 3.66 2 b
## Semé 3.06 0.308 35 2.434 3.69 2 b
##
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.3.2.2.4 Ab Poll
LM2_Ab_Po <- lmer(Ab_Poll ~ Area_gis_m_sq + Temperature + Periode * Gestion_moment_5 + (1|Parcelle), data = Inv)
# check_model(LM2_Ab_Po)
# shapiro.test(residuals(LM2_Ab_Po))
LM2_Ab_Po <- lmer(sqrt(Ab_Poll) ~ Area_gis_m_sq + Temperature + Periode * Gestion_moment_5 + (1|Parcelle), data = Inv)
# check_model(LM2_Ab_Po)
# shapiro.test(residuals(LM2_Ab_Po))
step(LM2_Ab_Po, direction = "backward") # sqrt(Ab_Poll) ~ Periode + Gestion_moment_5 + (1 | Parcelle) + Periode:Gestion_moment_5check_model(LM2_Ab_Po)
Anova(LM2_Ab_Po)## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: sqrt(Ab_Poll)
## Chisq Df Pr(>Chisq)
## Area_gis_m_sq 0.1000 1 0.7518512
## Temperature 0.5316 1 0.4659370
## Periode 9.6825 2 0.0078974 **
## Gestion_moment_5 116.0088 4 < 2.2e-16 ***
## Periode:Gestion_moment_5 29.0253 8 0.0003139 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# summary(LM2_Ab_Po)
# LM2_Ab_Po
eLM2_Ab_Po_P<-emmeans(LM2_Ab_Po,"Periode")
mcLM2_Ab_Po_P<-cld(eLM2_Ab_Po_P,ajust="tukey")
mcLM2_Ab_Po_P$.group<-as.numeric(mcLM2_Ab_Po_P$.group)
mcLM2_Ab_Po_P$group[mcLM2_Ab_Po_P$.group == 1] <- "a"
mcLM2_Ab_Po_P$group[mcLM2_Ab_Po_P$.group == 2] <- "b"
mcLM2_Ab_Po_P## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 3.01 0.232 108.4 2.55 3.47 1 a
## Juin 3.13 0.208 98.8 2.72 3.54 1 a
## Mi-juillet 3.35 0.205 97.2 2.94 3.76 1 a
##
## Results are averaged over the levels of: Gestion_moment_5
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Ab_P_G <- Inv %>%
filter(Gestion_moment_5 == "Graminées")
LM2_S_Ab_P_G <- lmer(sqrt(Ab_Poll) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_S_Ab_P_G)
eLM2_S_Ab_P_G<-emmeans(LM2_S_Ab_P_G,"Periode")
mcLM2_S_Ab_P_G<-cld(eLM2_S_Ab_P_G,ajust="tukey")
mcLM2_S_Ab_P_G$.group<-as.numeric(mcLM2_S_Ab_P_G$.group)
mcLM2_S_Ab_P_G$group[mcLM2_S_Ab_P_G$.group == 1] <- "a"
mcLM2_S_Ab_P_G$group[mcLM2_S_Ab_P_G$.group == 2] <- "b"
mcLM2_S_Ab_P_G## Periode emmean SE df lower.CL upper.CL .group group
## Juin 1.13 1.06 5.85 -1.475 3.73 1 a
## Mi-juillet 2.37 1.03 5.87 -0.157 4.89 1 a
## Fin juillet 3.68 1.18 5.99 0.797 6.57 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Ab_P_F <- Inv %>%
filter(Gestion_moment_5 == "Fleuri")
LM2_S_Ab_P_F <- lmer(sqrt(Ab_Poll) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_S_Ab_P_F)
eLM2_S_Ab_P_F<-emmeans(LM2_S_Ab_P_F,"Periode")
mcLM2_S_Ab_P_F<-cld(eLM2_S_Ab_P_F,ajust="tukey")
mcLM2_S_Ab_P_F$.group<-as.numeric(mcLM2_S_Ab_P_F$.group)
mcLM2_S_Ab_P_F$group[mcLM2_S_Ab_P_F$.group == 1] <- "a"
mcLM2_S_Ab_P_F$group[mcLM2_S_Ab_P_F$.group == 2] <- "b"
mcLM2_S_Ab_P_F## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 3.33 0.471 18.5 2.34 4.31 1 a
## Mi-juillet 4.14 0.427 15.0 3.23 5.05 1 a
## Juin 4.43 0.429 15.2 3.52 5.34 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Ab_P_S <- Inv %>%
filter(Gestion_moment_5 == "Semé")
LM2_S_Ab_P_S <- lmer(sqrt(Ab_Poll) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_S_Ab_P_S)
eLM2_S_Ab_P_S<-emmeans(LM2_S_Ab_P_S,"Periode")
mcLM2_S_Ab_P_S<-cld(eLM2_S_Ab_P_S,ajust="tukey")
mcLM2_S_Ab_P_S$.group<-as.numeric(mcLM2_S_Ab_P_S$.group)
mcLM2_S_Ab_P_S$group[mcLM2_S_Ab_P_S$.group == 1] <- "a"
mcLM2_S_Ab_P_S$group[mcLM2_S_Ab_P_S$.group == 2] <- "b"
mcLM2_S_Ab_P_S$group[mcLM2_S_Ab_P_S$.group == 12] <- "ab"
mcLM2_S_Ab_P_S## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 4.73 0.384 11.1 3.88 5.57 1 a
## Juin 5.43 0.369 10.8 4.61 6.24 1 a
## Mi-juillet 6.95 0.371 10.8 6.13 7.76 2 b
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Ab_P_Tr <- Inv %>%
filter(Gestion_moment_5 == "Tonte récente")
LM2_S_Ab_P_Tr <- lmer(sqrt(Ab_Poll) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_S_Ab_P_Tr)
eLM2_S_Ab_P_Tr<-emmeans(LM2_S_Ab_P_Tr,"Periode")
mcLM2_S_Ab_P_Tr<-cld(eLM2_S_Ab_P_Tr,ajust="tukey")
mcLM2_S_Ab_P_Tr$.group<-as.numeric(mcLM2_S_Ab_P_Tr$.group)
mcLM2_S_Ab_P_Tr$group[mcLM2_S_Ab_P_Tr$.group == 1] <- "a"
mcLM2_S_Ab_P_Tr$group[mcLM2_S_Ab_P_Tr$.group == 2] <- "b"
mcLM2_S_Ab_P_Tr## Periode emmean SE df lower.CL upper.CL .group group
## Mi-juillet 0.92 0.218 21.0 0.465 1.37 1 a
## Fin juillet 1.33 0.394 19.8 0.506 2.15 1 a
## Juin 1.62 0.293 20.9 1.014 2.23 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_S_Ab_P_Tt <- Inv %>%
filter(Gestion_moment_5 == "Tonte tardive")
LM2_S_Ab_P_Tt <- lmer(sqrt(Ab_Poll) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM2_S_Ab_P_Tt)
eLM2_S_Ab_P_Tt<-emmeans(LM2_S_Ab_P_Tt,"Periode")
mcLM2_S_Ab_P_Tt<-cld(eLM2_S_Ab_P_Tt,ajust="tukey")
mcLM2_S_Ab_P_Tt$.group<-as.numeric(mcLM2_S_Ab_P_Tt$.group)
mcLM2_S_Ab_P_Tt$group[mcLM2_S_Ab_P_Tt$.group == 1] <- "a"
mcLM2_S_Ab_P_Tt$group[mcLM2_S_Ab_P_Tt$.group == 2] <- "b"
mcLM2_S_Ab_P_Tt$group[mcLM2_S_Ab_P_Tt$.group == 12] <- "ab"
mcLM2_S_Ab_P_Tt## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 1.85 0.251 43.2 1.34 2.35 1 a
## Mi-juillet 2.70 0.319 44.0 2.05 3.34 12 ab
## Juin 3.03 0.282 43.4 2.47 3.60 2 b
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.eLM2_Ab_Po_G5<-emmeans(LM2_Ab_Po,"Gestion_moment_5")
mcLM2_Ab_Po_G5<-cld(eLM2_Ab_Po_G5,ajust="tukey")
mcLM2_Ab_Po_G5$.group<-as.numeric(mcLM2_Ab_Po_G5$.group)
mcLM2_Ab_Po_G5$group[mcLM2_Ab_Po_G5$.group == 1] <- "a"
mcLM2_Ab_Po_G5$group[mcLM2_Ab_Po_G5$.group == 2] <- "b"
mcLM2_Ab_Po_G5$group[mcLM2_Ab_Po_G5$.group == 3] <- "c"
mcLM2_Ab_Po_G5$group[mcLM2_Ab_Po_G5$.group == 12] <- "ab"
mcLM2_Ab_Po_G5$group[mcLM2_Ab_Po_G5$.group == 4] <- "d"
mcLM2_Ab_Po_G5## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Tonte récente 1.23 0.278 96.5 0.679 1.78 1 a
## Graminées 2.36 0.479 42.4 1.393 3.32 12 ab
## Tonte tardive 2.53 0.214 58.4 2.100 2.96 2 b
## Fleuri 3.97 0.293 43.1 3.383 4.56 3 c
## Semé 5.73 0.384 38.5 4.952 6.50 4 d
##
## Results are averaged over the levels of: Periode
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_Ab_Po_G5_Ju <- Inv %>%
filter(Periode == "Juin")
LM2_Ab_Po_G5_Ju <- lm(sqrt(Ab_Poll) ~ Area_gis_m_sq + Temperature + Gestion_moment_5, data = CM2_Ab_Po_G5_Ju)
eLM2_Ab_Po_G5_Ju<-emmeans(LM2_Ab_Po_G5_Ju,"Gestion_moment_5")
mcLM2_Ab_Po_G5_Ju<-cld(eLM2_Ab_Po_G5_Ju,ajust="tukey")
mcLM2_Ab_Po_G5_Ju$.group<-as.numeric(mcLM2_Ab_Po_G5_Ju$.group)
mcLM2_Ab_Po_G5_Ju$group[mcLM2_Ab_Po_G5_Ju$.group == 1] <- "a"
mcLM2_Ab_Po_G5_Ju$group[mcLM2_Ab_Po_G5_Ju$.group == 2] <- "b"
mcLM2_Ab_Po_G5_Ju$group[mcLM2_Ab_Po_G5_Ju$.group == 12] <- "ab"
mcLM2_Ab_Po_G5_Ju$group[mcLM2_Ab_Po_G5_Ju$.group == 3] <- "c"
mcLM2_Ab_Po_G5_Ju## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Graminées 1.29 0.557 39 0.168 2.42 12 ab
## Tonte récente 1.44 0.408 39 0.616 2.26 1 a
## Tonte tardive 2.92 0.287 39 2.343 3.50 2 b
## Fleuri 4.50 0.341 39 3.806 5.19 3 c
## Semé 5.39 0.461 39 4.459 6.32 3 c
##
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_Ab_Po_G5_mJ <- Inv %>%
filter(Periode == "Mi-juillet")
LM2_Ab_Po_G5_mJ <- lm(sqrt(Ab_Poll) ~ Area_gis_m_sq + Temperature + Gestion_moment_5, data = CM2_Ab_Po_G5_mJ)
eLM2_Ab_Po_G5_mJ<-emmeans(LM2_Ab_Po_G5_mJ,"Gestion_moment_5")
mcLM2_Ab_Po_G5_mJ<-cld(eLM2_Ab_Po_G5_mJ,ajust="tukey")
mcLM2_Ab_Po_G5_mJ$.group<-as.numeric(mcLM2_Ab_Po_G5_mJ$.group)
mcLM2_Ab_Po_G5_mJ$group[mcLM2_Ab_Po_G5_mJ$.group == 1] <- "a"
mcLM2_Ab_Po_G5_mJ$group[mcLM2_Ab_Po_G5_mJ$.group == 12] <- "ab"
mcLM2_Ab_Po_G5_mJ$group[mcLM2_Ab_Po_G5_mJ$.group == 23] <- "bc"
mcLM2_Ab_Po_G5_mJ$group[mcLM2_Ab_Po_G5_mJ$.group == 3] <- "c"
mcLM2_Ab_Po_G5_mJ$group[mcLM2_Ab_Po_G5_mJ$.group == 4] <- "d"
mcLM2_Ab_Po_G5_mJ## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Tonte récente 0.932 0.341 39 0.241 1.62 1 a
## Graminées 1.627 0.631 39 0.351 2.90 12 ab
## Tonte tardive 2.929 0.365 39 2.190 3.67 23 bc
## Fleuri 4.168 0.381 39 3.399 4.94 3 c
## Semé 6.872 0.509 39 5.843 7.90 4 d
##
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM2_Ab_Po_G5_fJ <- Inv %>%
filter(Periode == "Fin juillet")
LM2_Ab_Po_G5_fJ <- lm(sqrt(Ab_Poll) ~ Area_gis_m_sq + Temperature + Gestion_moment_5, data = CM2_Ab_Po_G5_fJ)
eLM2_Ab_Po_G5_fJ<-emmeans(LM2_Ab_Po_G5_fJ,"Gestion_moment_5")
mcLM2_Ab_Po_G5_fJ<-cld(eLM2_Ab_Po_G5_fJ,ajust="tukey")
mcLM2_Ab_Po_G5_fJ$.group<-as.numeric(mcLM2_Ab_Po_G5_fJ$.group)
mcLM2_Ab_Po_G5_fJ$group[mcLM2_Ab_Po_G5_fJ$.group == 1] <- "a"
mcLM2_Ab_Po_G5_fJ$group[mcLM2_Ab_Po_G5_fJ$.group == 2] <- "b"
mcLM2_Ab_Po_G5_fJ$group[mcLM2_Ab_Po_G5_fJ$.group == 12] <- "ab"
mcLM2_Ab_Po_G5_fJ## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Tonte récente 1.26 0.585 35 0.0763 2.45 1 a
## Tonte tardive 1.86 0.289 35 1.2705 2.44 1 a
## Fleuri 3.18 0.459 35 2.2468 4.11 12 ab
## Graminées 4.18 0.739 35 2.6801 5.68 2 b
## Semé 5.03 0.542 35 3.9281 6.13 2 b
##
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.3.2.3 LMM 3. - N_interactions & Probabilité de visite selon la quantité de fleurs
proba d’être visité: +/- fct de la densité: N_int ~ qtté_plantes (par sp. de fleurs)
LM3_qtte <- lmer(N_Interactions ~ Qtte_Plantes + Gestion_moment_5 + Qtte_Plantes:Gestion_moment_5 + Area_gis_m_sq + Temperature + Periode + Periode:Gestion_moment_5 + (1|Parcelle), data = Interactions_Gestion)
# check_model(LM3_qtte)
# shapiro.test(residuals(LM3_qtte))
LM3_qtte <- lmer(sqrt(N_Interactions) ~ Qtte_Plantes + Gestion_moment_5 + Qtte_Plantes:Gestion_moment_5 + Area_gis_m_sq + Temperature + Periode + Periode:Gestion_moment_5 + (1|Parcelle), data = Interactions_Gestion)
# check_model(LM3_qtte)
# shapiro.test(residuals(LM3_qtte))
step(LM3_qtte, direction = "backward")
# sqrt(N_Interactions) ~ Qtte_Plantes + Gestion_moment_5 + Temperature + Periode + (1 | Parcelle) + Qtte_Plantes:Gestion_moment_5 + Gestion_moment_5:Periodecheck_model(LM3_qtte)
Anova(LM3_qtte)## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: sqrt(N_Interactions)
## Chisq Df Pr(>Chisq)
## Qtte_Plantes 270.3715 1 < 2.2e-16 ***
## Gestion_moment_5 4.9530 4 0.292151
## Area_gis_m_sq 2.0722 1 0.150008
## Temperature 31.3862 1 2.115e-08 ***
## Periode 29.6885 2 3.575e-07 ***
## Qtte_Plantes:Gestion_moment_5 55.1174 4 3.070e-11 ***
## Gestion_moment_5:Periode 20.2190 8 0.009538 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1#summary(LM3_qtte)
#LM3_qtte
eLM3_qtte_P<-emmeans(LM3_qtte,"Periode")
mcLM3_qtte_P<-cld(eLM3_qtte_P,ajust="tukey")
mcLM3_qtte_P$.group<-as.numeric(mcLM3_qtte_P$.group)
mcLM3_qtte_P$group[mcLM3_qtte_P$.group == 1] <- "a"
mcLM3_qtte_P$group[mcLM3_qtte_P$.group == 2] <- "b"
mcLM3_qtte_P## Periode emmean SE df lower.CL upper.CL .group group
## Juin 1.20 0.0776 165 1.05 1.35 1 a
## Fin juillet 1.47 0.1001 118 1.27 1.66 2 b
## Mi-juillet 1.52 0.0826 196 1.36 1.69 2 b
##
## Results are averaged over the levels of: Gestion_moment_5
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.eLM3_qtte_G5<-emmeans(LM3_qtte,"Gestion_moment_5")
mcLM3_qtte_G5<-cld(eLM3_qtte_G5,ajust="tukey")
mcLM3_qtte_G5$.group<-as.numeric(mcLM3_qtte_G5$.group)
mcLM3_qtte_G5$group[mcLM3_qtte_G5$.group == 1] <- "a"
mcLM3_qtte_G5$group[mcLM3_qtte_G5$.group == 2] <- "b"
mcLM3_qtte_G5$group[mcLM3_qtte_G5$.group == 3] <- "c"
mcLM3_qtte_G5$group[mcLM3_qtte_G5$.group == 12] <- "ab"
mcLM3_qtte_G5$group[mcLM3_qtte_G5$.group == 4] <- "d"
mcLM3_qtte_G5## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Fleuri 1.16 0.0804 33.4 0.995 1.32 1 a
## Tonte tardive 1.26 0.0649 58.4 1.129 1.39 1 a
## Tonte récente 1.36 0.1180 397.1 1.130 1.59 1 a
## Semé 1.37 0.1014 25.7 1.159 1.58 1 a
## Graminées 1.83 0.3260 136.0 1.189 2.48 1 a
##
## Results are averaged over the levels of: Periode
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM3_P_G <- Interactions_Gestion %>%
filter(Gestion_moment_5 == "Graminées")
LM3_P_G <- lmer(sqrt(N_Interactions) ~ Qtte_Plantes + Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM3_P_G)
eLM3_P_G<-emmeans(LM3_P_G,"Periode")
mcLM3_P_G<-cld(eLM3_P_G,ajust="tukey")
mcLM3_P_G$.group<-as.numeric(mcLM3_P_G$.group)
mcLM3_P_G$group[mcLM3_P_G$.group == 1] <- "a"
mcLM3_P_G$group[mcLM3_P_G$.group == 2] <- "b"
mcLM3_P_G$group[mcLM3_P_G$.group == 12] <- "ab"
mcLM3_P_G## Periode emmean SE df lower.CL upper.CL .group group
## Juin 0.81 0.178 18.01 0.435 1.18 1 a
## Fin juillet 1.23 0.135 8.42 0.919 1.53 12 ab
## Mi-juillet 1.25 0.164 15.85 0.908 1.60 2 b
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM3_P_F <- Interactions_Gestion %>%
filter(Gestion_moment_5 == "Fleuri")
LM3_P_F <- lmer(sqrt(N_Interactions) ~ Qtte_Plantes + Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM3_P_F)
eLM3_P_F<-emmeans(LM3_P_F,"Periode")
mcLM3_P_F<-cld(eLM3_P_F,ajust="tukey")
mcLM3_P_F$.group<-as.numeric(mcLM3_P_F$.group)
mcLM3_P_F$group[mcLM3_P_F$.group == 1] <- "a"
mcLM3_P_F$group[mcLM3_P_F$.group == 2] <- "b"
mcLM3_P_F$group[mcLM3_P_F$.group == 12] <- "ab"
mcLM3_P_F## Periode emmean SE df lower.CL upper.CL .group group
## Juin 1.03 0.0797 10.7 0.858 1.21 1 a
## Fin juillet 1.09 0.0954 20.9 0.890 1.29 12 ab
## Mi-juillet 1.22 0.0818 11.7 1.040 1.40 2 b
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM3_P_S <- Interactions_Gestion %>%
filter(Gestion_moment_5 == "Semé")
LM3_P_S <- lmer(sqrt(N_Interactions) ~ Qtte_Plantes + Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM3_P_S)
eLM3_P_S<-emmeans(LM3_P_S,"Periode")
mcLM3_P_S<-cld(eLM3_P_S,ajust="tukey")
mcLM3_P_S$.group<-as.numeric(mcLM3_P_S$.group)
mcLM3_P_S$group[mcLM3_P_S$.group == 1] <- "a"
mcLM3_P_S## Periode emmean SE df lower.CL upper.CL .group group
## Juin 1.22 0.0536 11.89 1.10 1.34 1 a
## Fin juillet 1.24 0.0594 16.66 1.11 1.36 1 a
## Mi-juillet 1.31 0.0479 7.38 1.20 1.43 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM3_P_Tr <- Interactions_Gestion %>%
filter(Gestion_moment_5 == "Tonte récente")
LM3_P_Tr <- lmer(sqrt(N_Interactions) ~ Qtte_Plantes + Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM3_P_Tr)
eLM3_P_Tr<-emmeans(LM3_P_Tr,"Periode")
mcLM3_P_Tr<-cld(eLM3_P_Tr,ajust="tukey")
mcLM3_P_Tr$.group<-as.numeric(mcLM3_P_Tr$.group)
mcLM3_P_Tr$group[mcLM3_P_Tr$.group == 1] <- "a"
mcLM3_P_Tr## Periode emmean SE df lower.CL upper.CL .group group
## Juin 0.756 0.216 19.7 0.304 1.21 1 a
## Mi-juillet 1.076 0.200 28.3 0.666 1.49 1 a
## Fin juillet 1.307 0.284 38.3 0.733 1.88 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM3_P_Tt <- Interactions_Gestion %>%
filter(Gestion_moment_5 == "Tonte tardive")
LM3_P_Tt <- lmer(sqrt(N_Interactions) ~ Qtte_Plantes + Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM3_P_Tt)
eLM3_P_Tt<-emmeans(LM3_P_Tt,"Periode")
mcLM3_P_Tt<-cld(eLM3_P_Tt,ajust="tukey")
mcLM3_P_Tt$.group<-as.numeric(mcLM3_P_Tt$.group)
mcLM3_P_Tt$group[mcLM3_P_Tt$.group == 1] <- "a"
mcLM3_P_Tt$group[mcLM3_P_Tt$.group == 2] <- "b"
mcLM3_P_Tt$group[mcLM3_P_Tt$.group == 12] <- "ab"
mcLM3_P_Tt## Periode emmean SE df lower.CL upper.CL .group group
## Juin 1.26 0.112 40.2 1.03 1.49 1 a
## Fin juillet 1.50 0.116 68.1 1.27 1.74 12 ab
## Mi-juillet 1.75 0.123 50.9 1.51 2.00 2 b
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM3_G5_Ju <- Interactions_Gestion %>%
filter(Periode == "Juin")
LM3_G5_Ju <- lmer(sqrt(N_Interactions) ~ Qtte_Plantes + Area_gis_m_sq + Temperature + Gestion_moment_5 + (1|Parcelle), data = CM3_G5_Ju)
eLM3_G5_Ju<-emmeans(LM3_G5_Ju,"Gestion_moment_5")
mcLM3_G5_Ju<-cld(eLM3_G5_Ju,ajust="tukey")
mcLM3_G5_Ju$.group<-as.numeric(mcLM3_G5_Ju$.group)
mcLM3_G5_Ju$group[mcLM3_G5_Ju$.group == 1] <- "a"
mcLM3_G5_Ju$group[mcLM3_G5_Ju$.group == 2] <- "b"
mcLM3_G5_Ju## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Graminées 0.909 0.274 38.0 0.353 1.46 1 a
## Tonte récente 1.082 0.179 77.9 0.725 1.44 1 a
## Fleuri 1.087 0.106 25.8 0.869 1.31 1 a
## Tonte tardive 1.216 0.108 40.5 0.998 1.43 1 a
## Semé 1.248 0.140 23.0 0.959 1.54 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM3_G5_mJ <- Interactions_Gestion %>%
filter(Periode == "Mi-juillet")
LM3_G5_mJ <- lmer(sqrt(N_Interactions) ~ Qtte_Plantes + Area_gis_m_sq + Temperature + Gestion_moment_5 + (1|Parcelle), data = CM3_G5_mJ)
eLM3_G5_mJ<-emmeans(LM3_G5_mJ,"Gestion_moment_5")
mcLM3_G5_mJ<-cld(eLM3_G5_mJ,ajust="tukey")
mcLM3_G5_mJ$.group<-as.numeric(mcLM3_G5_mJ$.group)
mcLM3_G5_mJ$group[mcLM3_G5_mJ$.group == 1] <- "a"
mcLM3_G5_mJ$group[mcLM3_G5_mJ$.group == 2] <- "b"
mcLM3_G5_mJ$group[mcLM3_G5_mJ$.group == 3] <- "c"
mcLM3_G5_mJ## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Tonte récente 1.02 0.163 166.2 0.701 1.34 1 a
## Fleuri 1.30 0.100 22.6 1.097 1.51 1 a
## Graminées 1.31 0.241 31.2 0.815 1.80 1 a
## Tonte tardive 1.36 0.113 39.4 1.131 1.59 1 a
## Semé 1.42 0.118 16.6 1.172 1.67 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM3_G5_fJ <- Interactions_Gestion %>%
filter(Periode == "Fin juillet")
LM3_G5_fJ <- lmer(sqrt(N_Interactions) ~ Qtte_Plantes + Area_gis_m_sq + Temperature + Gestion_moment_5 + (1|Parcelle), data = CM3_G5_fJ)
eLM3_G5_fJ<-emmeans(LM3_G5_fJ,"Gestion_moment_5")
mcLM3_G5_fJ<-cld(eLM3_G5_fJ,ajust="tukey")
mcLM3_G5_fJ$.group<-as.numeric(mcLM3_G5_fJ$.group)
mcLM3_G5_fJ$group[mcLM3_G5_fJ$.group == 1] <- "a"
mcLM3_G5_fJ$group[mcLM3_G5_fJ$.group == 2] <- "b"
mcLM3_G5_fJ$group[mcLM3_G5_fJ$.group == 12] <- "ab"
mcLM3_G5_fJ## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Tonte tardive 0.928 0.0893 47.8 0.748 1.11 1 a
## Tonte récente 1.097 0.2219 74.5 0.655 1.54 1 a
## Fleuri 1.219 0.1183 22.8 0.975 1.46 1 a
## Semé 1.388 0.1244 15.7 1.124 1.65 1 a
## Graminées 1.493 0.1883 21.5 1.102 1.88 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.3.2.4 LMM 4. - Sum(N_interactions)
Int_glm <- Interactions_Gestion %>%
group_by(Site_gestion_date, Periode, Gestion_moment_5, Area_gis_m_sq, Temperature, Parcelle) %>%
summarize(n= sum(N_Interactions))
LM4_Int <- lmer(n ~ Area_gis_m_sq + Temperature + Periode * Gestion_moment_5 + (1|Parcelle), data = Int_glm)
# check_model(LM4_Int)
# shapiro.test(residuals(LM4_Int))
LM4_Int <- lmer(sqrt(n) ~ Area_gis_m_sq + Temperature + Periode * Gestion_moment_5 + (1|Parcelle), data = Int_glm)
# check_model(LM4_Int)
# shapiro.test(residuals(LM4_Int))
step(LM4_Int, direction = "backward")
# sqrt(n) ~ Temperature + Periode + Gestion_moment_5 + (1 | Parcelle) + Periode:Gestion_moment_5check_model(LM4_Int)
Anova(LM4_Int)## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: sqrt(n)
## Chisq Df Pr(>Chisq)
## Area_gis_m_sq 2.1736 1 0.140394
## Temperature 5.0618 1 0.024459 *
## Periode 9.7773 2 0.007532 **
## Gestion_moment_5 73.7697 4 3.627e-15 ***
## Periode:Gestion_moment_5 16.3526 8 0.037603 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1#summary(LM4_Int)
#LM4_Int
eLM4_Int_P<-emmeans(LM4_Int,"Periode")
mcLM4_Int_P<-cld(eLM4_Int_P,ajust="tukey")
mcLM4_Int_P$.group<-as.numeric(mcLM4_Int_P$.group)
mcLM4_Int_P$group[mcLM4_Int_P$.group == 1] <- "a"
mcLM4_Int_P$group[mcLM4_Int_P$.group == 2] <- "b"
mcLM4_Int_P## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 4.53 0.510 112 3.52 5.54 1 a
## Juin 4.73 0.455 105 3.83 5.63 1 a
## Mi-juillet 5.52 0.448 104 4.63 6.41 1 a
##
## Results are averaged over the levels of: Gestion_moment_5
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM4_P_G <- Int_glm %>%
filter(Gestion_moment_5 == "Graminées")
LM4_P_G <- lmer(sqrt(n) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM4_P_G)
eLM4_P_G<-emmeans(LM4_P_G,"Periode")
mcLM4_P_G<-cld(eLM4_P_G,ajust="tukey")
mcLM4_P_G$.group<-as.numeric(mcLM4_P_G$.group)
mcLM4_P_G$group[mcLM4_P_G$.group == 1] <- "a"
mcLM4_P_G$group[mcLM4_P_G$.group == 2] <- "b"
mcLM4_P_G## Periode emmean SE df lower.CL upper.CL .group group
## Juin 1.06 1.58 5.85 -2.835 4.95 1 a
## Mi-juillet 2.71 1.53 5.87 -1.067 6.48 1 a
## Fin juillet 4.91 1.76 5.99 0.589 9.22 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM4_P_F <- Int_glm %>%
filter(Gestion_moment_5 == "Fleuri")
LM4_P_F <- lmer(sqrt(n) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM4_P_F)
eLM4_P_F<-emmeans(LM4_P_F,"Periode")
mcLM4_P_F<-cld(eLM4_P_F,ajust="tukey")
mcLM4_P_F$.group<-as.numeric(mcLM4_P_F$.group)
mcLM4_P_F$group[mcLM4_P_F$.group == 1] <- "a"
mcLM4_P_F$group[mcLM4_P_F$.group == 2] <- "b"
mcLM4_P_F$group[mcLM4_P_F$.group == 12] <- "ab"
mcLM4_P_F## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 4.05 0.889 19.8 2.19 5.90 1 a
## Juin 5.91 0.799 16.3 4.21 7.60 12 ab
## Mi-juillet 6.30 0.796 16.2 4.62 7.99 2 b
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM4_P_S <- Int_glm %>%
filter(Gestion_moment_5 == "Semé")
LM4_P_S <- lmer(sqrt(n) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM4_P_S)
eLM4_P_S<-emmeans(LM4_P_S,"Periode")
mcLM4_P_S<-cld(eLM4_P_S,ajust="tukey")
mcLM4_P_S$.group<-as.numeric(mcLM4_P_S$.group)
mcLM4_P_S$group[mcLM4_P_S$.group == 1] <- "a"
mcLM4_P_S$group[mcLM4_P_S$.group == 2] <- "b"
mcLM4_P_S$group[mcLM4_P_S$.group == 12] <- "ab"
mcLM4_P_S## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 7.39 0.977 8.37 5.16 9.63 1 a
## Juin 8.42 0.950 7.91 6.22 10.61 1 a
## Mi-juillet 11.30 0.954 7.98 9.10 13.50 2 b
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM4_P_Tr <- Int_glm %>%
filter(Gestion_moment_5 == "Tonte récente")
LM4_P_Tr <- lmer(sqrt(n) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM4_P_Tr)
eLM4_P_Tr<-emmeans(LM4_P_Tr,"Periode")
mcLM4_P_Tr<-cld(eLM4_P_Tr,ajust="tukey")
mcLM4_P_Tr$.group<-as.numeric(mcLM4_P_Tr$.group)
mcLM4_P_Tr$group[mcLM4_P_Tr$.group == 1] <- "a"
mcLM4_P_Tr$group[mcLM4_P_Tr$.group == 2] <- "b"
mcLM4_P_Tr## Periode emmean SE df lower.CL upper.CL .group group
## Mi-juillet 1.26 0.517 21.0 0.179 2.33 1 a
## Juin 2.57 0.693 20.9 1.123 4.01 1 a
## Fin juillet 2.61 0.934 19.8 0.664 4.56 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM4_P_Tt <- Int_glm %>%
filter(Gestion_moment_5 == "Tonte tardive")
LM4_P_Tt <- lmer(sqrt(n) ~ Area_gis_m_sq + Temperature + Periode + (1|Parcelle), data = CM4_P_Tt)
eLM4_P_Tt<-emmeans(LM4_P_Tt,"Periode")
mcLM4_P_Tt<-cld(eLM4_P_Tt,ajust="tukey")
mcLM4_P_Tt$.group<-as.numeric(mcLM4_P_Tt$.group)
mcLM4_P_Tt$group[mcLM4_P_Tt$.group == 1] <- "a"
mcLM4_P_Tt$group[mcLM4_P_Tt$.group == 2] <- "b"
mcLM4_P_Tt$group[mcLM4_P_Tt$.group == 12] <- "ab"
mcLM4_P_Tt## Periode emmean SE df lower.CL upper.CL .group group
## Fin juillet 3.44 0.700 43.4 2.03 4.85 1 a
## Juin 5.72 0.788 43.4 4.13 7.31 1 a
## Mi-juillet 6.03 0.892 44.0 4.23 7.83 1 a
##
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.eLM4_Int_G5<-emmeans(LM4_Int,"Gestion_moment_5")
mcLM4_Int_G5<-cld(eLM4_Int_G5,ajust="tukey")
mcLM4_Int_G5$.group<-as.numeric(mcLM4_Int_G5$.group)
mcLM4_Int_G5$group[mcLM4_Int_G5$.group == 1] <- "a"
mcLM4_Int_G5$group[mcLM4_Int_G5$.group == 2] <- "b"
mcLM4_Int_G5$group[mcLM4_Int_G5$.group == 3] <- "c"
mcLM4_Int_G5$group[mcLM4_Int_G5$.group == 12] <- "ab"
mcLM4_Int_G5$group[mcLM4_Int_G5$.group == 4] <- "d"
mcLM4_Int_G5## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Tonte récente 1.89 0.606 99.0 0.683 3.09 1 a
## Graminées 3.02 1.002 42.5 1.003 5.05 12 ab
## Tonte tardive 5.02 0.456 60.9 4.112 5.93 2 b
## Fleuri 5.40 0.614 43.4 4.163 6.64 2 b
## Semé 9.29 0.799 38.1 7.672 10.91 3 c
##
## Results are averaged over the levels of: Periode
## Degrees-of-freedom method: kenward-roger
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM4_G5_Ju <- Int_glm %>%
filter(Periode == "Juin")
LM4_G5_Ju <- lm(sqrt(n) ~ Area_gis_m_sq + Temperature + Gestion_moment_5, data = CM4_G5_Ju)
eLM4_G5_Ju<-emmeans(LM4_G5_Ju,"Gestion_moment_5")
mcLM4_G5_Ju<-cld(eLM4_G5_Ju,ajust="tukey")
mcLM4_G5_Ju$.group<-as.numeric(mcLM4_G5_Ju$.group)
mcLM4_G5_Ju$group[mcLM4_G5_Ju$.group == 1] <- "a"
mcLM4_G5_Ju$group[mcLM4_G5_Ju$.group == 2] <- "b"
mcLM4_G5_Ju## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Graminées 1.28 1.193 39 -1.133 3.69 1 a
## Tonte récente 1.96 0.873 39 0.189 3.72 1 a
## Tonte tardive 5.64 0.615 39 4.398 6.89 2 b
## Fleuri 5.73 0.730 39 4.252 7.21 2 b
## Semé 8.26 0.987 39 6.260 10.25 2 b
##
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM4_G5_mJ <- Int_glm %>%
filter(Periode == "Mi-juillet")
LM4_G5_mJ <- lm(sqrt(n) ~ Area_gis_m_sq + Temperature + Gestion_moment_5, data = CM4_G5_mJ)
eLM4_G5_mJ<-emmeans(LM4_G5_mJ,"Gestion_moment_5")
mcLM4_G5_mJ<-cld(eLM4_G5_mJ,ajust="tukey")
mcLM4_G5_mJ$.group<-as.numeric(mcLM4_G5_mJ$.group)
mcLM4_G5_mJ$group[mcLM4_G5_mJ$.group == 1] <- "a"
mcLM4_G5_mJ$group[mcLM4_G5_mJ$.group == 2] <- "b"
mcLM4_G5_mJ$group[mcLM4_G5_mJ$.group == 3] <- "c"
mcLM4_G5_mJ## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Tonte récente 1.28 0.731 39 -0.196 2.76 1 a
## Graminées 1.71 1.352 39 -1.026 4.44 1 a
## Fleuri 6.46 0.815 39 4.812 8.11 2 b
## Tonte tardive 6.55 0.783 39 4.966 8.13 2 b
## Semé 11.29 1.090 39 9.088 13.50 3 c
##
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM4_G5_fJ <- Int_glm %>%
filter(Periode == "Fin juillet")
LM4_G5_fJ <- lm(sqrt(n) ~ Area_gis_m_sq + Temperature + Gestion_moment_5, data = CM4_G5_fJ)
eLM4_G5_fJ<-emmeans(LM4_G5_fJ,"Gestion_moment_5")
mcLM4_G5_fJ<-cld(eLM4_G5_fJ,ajust="tukey")
mcLM4_G5_fJ$.group<-as.numeric(mcLM4_G5_fJ$.group)
mcLM4_G5_fJ$group[mcLM4_G5_fJ$.group == 1] <- "a"
mcLM4_G5_fJ$group[mcLM4_G5_fJ$.group == 2] <- "b"
mcLM4_G5_fJ$group[mcLM4_G5_fJ$.group == 12] <- "ab"
mcLM4_G5_fJ## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Tonte récente 1.82 1.247 35 -0.71 4.35 1 a
## Tonte tardive 3.40 0.615 35 2.15 4.65 1 a
## Fleuri 4.08 0.977 35 2.10 6.07 12 ab
## Graminées 5.69 1.576 35 2.49 8.89 12 ab
## Semé 8.30 1.155 35 5.95 10.64 2 b
##
## Results are given on the sqrt (not the response) scale.
## Confidence level used: 0.95
## Note: contrasts are still on the sqrt scale
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.3.2.5 LMM 5. Nombre de visites par unité florale par heure - nvis/fl.h
nvis./fl. h: Q où sp présente dans les 3 Q, sp phare, comparer période/gestion
Interactions_Gestion %>%
count(Sp_Pollinisateurs) %>%
arrange(desc(n))## Sp_Pollinisateurs n
## 1 Apis mellifera 419
## 2 318
## 3 Lasioglossum spec. 149
## 4 Bombus pascuorum 110
## 5 Hylaeus spec. 103
## 6 Bombus lapidarius 101
## 7 Syrphus spec. 87
## 8 Episyrphus balteatus 73
## 9 Anthophila indet. 71
## 10 Syritta pipiens 49
## 11 Eristalis tenax 39
## 12 Bombus vestalis/bohemicus 36
## 13 Syrphidae indet. 35
## 14 Sphaerophoria spec. 34
## 15 Sphaerophoria scripta 31
## 16 Bombus terrestris/lucorum/magnus/cryptarum 25
## 17 Heriades truncorum 25
## 18 Platycheirus albimanus 23
## 19 Maniola jurtina 22
## 20 Halictus scabiosae 21
## 21 Eupeodes corollae 19
## 22 Lasioglossum (Leuchalictus) - spec. 17
## 23 Polyommatus icarus 17
## 24 Andrena flavipes 16
## 25 Lasioglossum (Dialictus) - spec. 15
## 26 Eristalis spec. 14
## 27 Megachile spec. 14
## 28 Eristalis nemorum 12
## 29 Eristalis pertinax 11
## 30 Syrphus ribesii 11
## 31 Andrena rosae 9
## 32 Andrena spec. 9
## 33 Sphecodes spec. 8
## 34 Cheilosia spec. 7
## 35 Andrena fulvago 6
## 36 Helophilus pendulus 6
## 37 Osmia leaiana/niveata 6
## 38 Aricia agestis 4
## 39 Bombus pratorum 4
## 40 Eristalis arbustorum 4
## 41 Myathropa florea 4
## 42 Pieris spec. 4
## 43 Bombus lucorum 3
## 44 Ceratina cyanea 3
## 45 Chelostoma rapunculi 3
## 46 Dasypoda hirtipes 3
## 47 Eupeodes luniger 3
## 48 Lasioglossum (Sphecodogastra) - spec. 3
## 49 Melanostoma mellinum 3
## 50 Melanostoma mellinum - Platycheirus albimanus 3
## 51 Andrena humilis 2
## 52 Cerceris arenaria 2
## 53 Cerceris rybyensis 2
## 54 Cerceris spec. 2
## 55 Colletes daviesanus/fodiens/similis 2
## 56 Euclidia glyphica 2
## 57 Halictidae indet. 2
## 58 Helophilus trivittatus 2
## 59 Megachile centuncularis 2
## 60 Panurgus calcaratus 2
## 61 Pyronia tithonus 2
## 62 Aglais io 1
## 63 Andrena dorsata 1
## 64 Andrena minutula-gr. 1
## 65 Andrena proxima 1
## 66 Bombus spec. 1
## 67 Bombus vestalis 1
## 68 Carcharodus alceae 1
## 69 Cheilosia pagana 1
## 70 Cheilosia variabilis 1
## 71 Dasysyrphus albostriatus 1
## 72 Eristalis arbustorum/abusiva 1
## 73 Eupeodes latifasciatus 1
## 74 Hylaeus communis 1
## 75 Hylaeus gredleri 1
## 76 Lasioglossum leucozonium 1
## 77 Lasioglossum sexnotatum 1
## 78 Lycaena phlaeas 1
## 79 Macroglossum stellatarum 1
## 80 Megachile willughbiella 1
## 81 Melanostoma spec. 1
## 82 Mesembrina meridiana 1
## 83 Neoascia spec. 1
## 84 Nowickia ferox 1
## 85 Osmia spinulosa 1
## 86 Paragus haemorrhous 1
## 87 Pieris napi 1
## 88 Pieris rapae 1
## 89 Pipizella spec. 1
## 90 Pipizella viduata 1
## 91 Scaeva pyrastri 1
## 92 Scaeva selenitica 1
## 93 Scaeva spec. 1
## 94 Stelis punctulatissima 1
## 95 Stomorhina lunata 1Interactions_nvis_fl_h <- Interactions_Gestion %>%
filter(Sp_Pollinisateurs == "Apis mellifera"|
Sp_Pollinisateurs == "Bombus pascuorum"|
Sp_Pollinisateurs == "Bombus lapidarius"|
Sp_Pollinisateurs == "Lasioglossum spec."|
Sp_Pollinisateurs == "Hylaeus spec."|
Sp_Pollinisateurs == "Episyrphus balteatus")
Q1 <- Interactions_nvis_fl_h %>%
filter(Nombre_quadrats == "1") %>%
mutate(nvis_fl_h = N_Interactions/(Qtte_Plantes/2))
Q2 <- Interactions_nvis_fl_h %>%
filter(Nombre_quadrats == "2") %>%
mutate(nvis_fl_h = N_Interactions/(Qtte_Plantes/4))
Q3 <- Interactions_nvis_fl_h %>%
filter(Nombre_quadrats == "3") %>%
mutate(nvis_fl_h = N_Interactions/(Qtte_Plantes/6))
Interactions_nvis_fl_h <- full_join(Q1, Q2)
Interactions_nvis_fl_h <- full_join(Interactions_nvis_fl_h, Q3)
Interactions_nvis_fl_h %>% ggplot(aes (x = Sp_Pollinisateurs, y = nvis_fl_h, color = Gestion_moment_5)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(#title = "",
x = "Sp_poll", y = "nvis_fl_h") +
theme(legend.position = "bottom") +
Interactions_nvis_fl_h %>% ggplot(aes (x = Sp_Pollinisateurs, y = nvis_fl_h, color = Periode)) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(#title = "",
x = "Sp_poll", y = "nvis_fl_h") +
theme(legend.position = "bottom")
LM5_nvis <- lmer(nvis_fl_h ~ Area_gis_m_sq + Temperature + Periode * Gestion_moment_5 + Sp_Pollinisateurs + (1|Parcelle), data = Interactions_nvis_fl_h)
# check_model(LM5_nvis)
# shapiro.test(residuals(LM5_nvis))
LM5_nvis <- lmer(log(nvis_fl_h) ~ Area_gis_m_sq + Temperature + Periode * Gestion_moment_5 + Sp_Pollinisateurs + (1|Parcelle), data = Interactions_nvis_fl_h)
# check_model(LM5_nvis)
# shapiro.test(residuals(LM5_nvis))
step(LM5_nvis, direction = "backward")
# log(nvis_fl_h) ~ log(nvis_fl_h) ~ Temperature + Periode + Gestion_moment_5 + Sp_Pollinisateurs + (1 | Parcelle) + Periode:Gestion_moment_5check_model(LM5_nvis)
Anova(LM5_nvis)## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: log(nvis_fl_h)
## Chisq Df Pr(>Chisq)
## Area_gis_m_sq 0.5189 1 0.471318
## Temperature 24.5067 1 7.405e-07 ***
## Periode 16.7118 2 0.000235 ***
## Gestion_moment_5 56.4344 4 1.626e-11 ***
## Sp_Pollinisateurs 15.0319 5 0.010227 *
## Periode:Gestion_moment_5 19.4051 8 0.012837 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1#summary(LM5_nvis)
#LM5_nvis
eLM5_nvis_P<-emmeans(LM5_nvis,"Periode")
mcLM5_nvis_P<-cld(eLM5_nvis_P,ajust="tukey")
mcLM5_nvis_P$.group<-as.numeric(mcLM5_nvis_P$.group)
mcLM5_nvis_P$group[mcLM5_nvis_P$.group == 1] <- "a"
mcLM5_nvis_P$group[mcLM5_nvis_P$.group == 2] <- "b"
mcLM5_nvis_P## Periode emmean SE df lower.CL upper.CL .group group
## Juin -1.501 0.161 178 -1.82 -1.182 1 a
## Mi-juillet -1.021 0.141 124 -1.30 -0.742 2 b
## Fin juillet -0.899 0.145 126 -1.19 -0.611 2 b
##
## Results are averaged over the levels of: Gestion_moment_5, Sp_Pollinisateurs
## Degrees-of-freedom method: kenward-roger
## Results are given on the log (not the response) scale.
## Confidence level used: 0.95
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM5_P_G <- Interactions_nvis_fl_h %>%
filter(Gestion_moment_5 == "Graminées")
LM5_P_G <- lmer(log(nvis_fl_h) ~ Area_gis_m_sq + Temperature + Periode + Sp_Pollinisateurs + (1|Parcelle), data = CM5_P_G)
eLM5_P_G<-emmeans(LM5_P_G,"Periode")
mcLM5_P_G<-cld(eLM5_P_G,ajust="tukey")
mcLM5_P_G$.group<-as.numeric(mcLM5_P_G$.group)
mcLM5_P_G$group[mcLM5_P_G$.group == 1] <- "a"
mcLM5_P_G$group[mcLM5_P_G$.group == 2] <- "b"
mcLM5_P_G$group[mcLM5_P_G$.group == 12] <- "ab"
mcLM5_P_G## Periode emmean SE df lower.CL upper.CL .group group
## Juin -3.274 1.316 78.1 -5.89 -0.654 12 ab
## Mi-juillet -2.069 0.583 36.6 -3.25 -0.888 1 a
## Fin juillet -0.028 0.635 27.9 -1.33 1.273 2 b
##
## Results are averaged over the levels of: Sp_Pollinisateurs
## Degrees-of-freedom method: kenward-roger
## Results are given on the log (not the response) scale.
## Confidence level used: 0.95
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM5_P_F <- Interactions_nvis_fl_h %>%
filter(Gestion_moment_5 == "Fleuri")
LM5_P_F <- lmer(log(nvis_fl_h) ~ Area_gis_m_sq + Temperature + Periode + Sp_Pollinisateurs + (1|Parcelle), data = CM5_P_F)
eLM5_P_F<-emmeans(LM5_P_F,"Periode")
mcLM5_P_F<-cld(eLM5_P_F,ajust="tukey")
mcLM5_P_F$.group<-as.numeric(mcLM5_P_F$.group)
mcLM5_P_F$group[mcLM5_P_F$.group == 1] <- "a"
mcLM5_P_F$group[mcLM5_P_F$.group == 2] <- "b"
mcLM5_P_F$group[mcLM5_P_F$.group == 12] <- "ab"
mcLM5_P_F## Periode emmean SE df lower.CL upper.CL .group group
## Juin -0.954 0.145 19.5 -1.26 -0.651 1 a
## Mi-juillet -0.774 0.148 20.4 -1.08 -0.466 1 a
## Fin juillet -0.622 0.228 69.3 -1.08 -0.166 1 a
##
## Results are averaged over the levels of: Sp_Pollinisateurs
## Degrees-of-freedom method: kenward-roger
## Results are given on the log (not the response) scale.
## Confidence level used: 0.95
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM5_P_S <- Interactions_nvis_fl_h %>%
filter(Gestion_moment_5 == "Semé")
LM5_P_S <- lmer(log(nvis_fl_h) ~ Area_gis_m_sq + Temperature + Periode + Sp_Pollinisateurs + (1|Parcelle), data = CM5_P_S)
eLM5_P_S<-emmeans(LM5_P_S,"Periode")
mcLM5_P_S<-cld(eLM5_P_S,ajust="tukey")
mcLM5_P_S$.group<-as.numeric(mcLM5_P_S$.group)
mcLM5_P_S$group[mcLM5_P_S$.group == 1] <- "a"
mcLM5_P_S$group[mcLM5_P_S$.group == 2] <- "b"
mcLM5_P_S$group[mcLM5_P_S$.group == 12] <- "ab"
mcLM5_P_S## Periode emmean SE df lower.CL upper.CL .group group
## Mi-juillet -0.273 0.149 5.27 -0.651 0.106 1 a
## Juin -0.154 0.159 6.66 -0.534 0.227 1 a
## Fin juillet -0.128 0.175 9.56 -0.521 0.265 1 a
##
## Results are averaged over the levels of: Sp_Pollinisateurs
## Degrees-of-freedom method: kenward-roger
## Results are given on the log (not the response) scale.
## Confidence level used: 0.95
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM5_P_Tr <- Interactions_nvis_fl_h %>%
filter(Gestion_moment_5 == "Tonte récente")
LM5_P_Tr <- lmer(log(nvis_fl_h) ~ Area_gis_m_sq + Temperature + Periode + Sp_Pollinisateurs + (1|Parcelle), data = CM5_P_Tr)
eLM5_P_Tr<-emmeans(LM5_P_Tr,"Periode")
mcLM5_P_Tr<-cld(eLM5_P_Tr,ajust="tukey")
mcLM5_P_Tr$.group<-as.numeric(mcLM5_P_Tr$.group)
mcLM5_P_Tr$group[mcLM5_P_Tr$.group == 1] <- "a"
mcLM5_P_Tr$group[mcLM5_P_Tr$.group == 2] <- "b"
mcLM5_P_Tr$group[mcLM5_P_Tr$.group == 12] <- "ab"
mcLM5_P_Tr## Periode emmean SE df lower.CL upper.CL .group group
## Juin -2.42 0.308 10.90 -3.10 -1.745 1 a
## Fin juillet -1.57 0.449 5.92 -2.67 -0.466 1 a
## Mi-juillet -1.34 0.365 12.72 -2.13 -0.545 1 a
##
## Results are averaged over the levels of: Sp_Pollinisateurs
## Degrees-of-freedom method: kenward-roger
## Results are given on the log (not the response) scale.
## Confidence level used: 0.95
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM5_P_Tt <- Interactions_nvis_fl_h %>%
filter(Gestion_moment_5 == "Tonte tardive")
LM5_P_Tt <- lmer(log(nvis_fl_h) ~ Area_gis_m_sq + Temperature + Periode + Sp_Pollinisateurs + (1|Parcelle), data = CM5_P_Tt)
eLM5_P_Tt<-emmeans(LM5_P_Tt,"Periode")
mcLM5_P_Tt<-cld(eLM5_P_Tt,ajust="tukey")
mcLM5_P_Tt$.group<-as.numeric(mcLM5_P_Tt$.group)
mcLM5_P_Tt$group[mcLM5_P_Tt$.group == 1] <- "a"
mcLM5_P_Tt$group[mcLM5_P_Tt$.group == 2] <- "b"
mcLM5_P_Tt$group[mcLM5_P_Tt$.group == 12] <- "ab"
mcLM5_P_Tt## Periode emmean SE df lower.CL upper.CL .group group
## Juin -2.24 0.171 49.7 -2.58 -1.894 1 a
## Mi-juillet -1.47 0.178 57.6 -1.83 -1.116 2 b
## Fin juillet -1.28 0.175 74.9 -1.62 -0.927 2 b
##
## Results are averaged over the levels of: Sp_Pollinisateurs
## Degrees-of-freedom method: kenward-roger
## Results are given on the log (not the response) scale.
## Confidence level used: 0.95
## P value adjustment: tukey method for comparing a family of 3 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.eLM5_nvis_G5<-emmeans(LM5_nvis,"Gestion_moment_5")
mcLM5_nvis_G5<-cld(eLM5_nvis_G5,ajust="tukey")
mcLM5_nvis_G5$.group<-as.numeric(mcLM5_nvis_G5$.group)
mcLM5_nvis_G5$group[mcLM5_nvis_G5$.group == 1] <- "a"
mcLM5_nvis_G5$group[mcLM5_nvis_G5$.group == 12] <- "ab"
mcLM5_nvis_G5$group[mcLM5_nvis_G5$.group == 23] <- "bc"
mcLM5_nvis_G5$group[mcLM5_nvis_G5$.group == 3] <- "c"
mcLM5_nvis_G5## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Tonte tardive -1.803 0.126 56.6 -2.056 -1.549 1 a
## Graminées -1.542 0.395 60.7 -2.331 -0.752 12 ab
## Tonte récente -1.408 0.217 251.4 -1.835 -0.980 12 ab
## Fleuri -0.751 0.161 36.1 -1.077 -0.426 23 bc
## Semé -0.198 0.196 24.7 -0.601 0.206 3 c
##
## Results are averaged over the levels of: Periode, Sp_Pollinisateurs
## Degrees-of-freedom method: kenward-roger
## Results are given on the log (not the response) scale.
## Confidence level used: 0.95
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM5_G5_Ju <- Interactions_nvis_fl_h %>%
filter(Periode == "Juin")
LM5_G5_Ju <- lmer(log(nvis_fl_h) ~ Area_gis_m_sq + Temperature + Gestion_moment_5 + Sp_Pollinisateurs + (1|Parcelle), data = CM5_G5_Ju)
eLM5_G5_Ju<-emmeans(LM5_G5_Ju,"Gestion_moment_5")
mcLM5_G5_Ju<-cld(eLM5_G5_Ju,ajust="tukey")
mcLM5_G5_Ju$.group<-as.numeric(mcLM5_G5_Ju$.group)
mcLM5_G5_Ju$group[mcLM5_G5_Ju$.group == 1] <- "a"
mcLM5_G5_Ju$group[mcLM5_G5_Ju$.group == 2] <- "b"
mcLM5_G5_Ju$group[mcLM5_G5_Ju$.group == 12] <- "ab"
mcLM5_G5_Ju## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Tonte récente -2.19 0.458 76.2 -3.10 -1.2737 1 a
## Tonte tardive -1.99 0.265 35.4 -2.52 -1.4496 1 a
## Graminées -1.85 0.930 38.1 -3.74 0.0287 12 ab
## Fleuri -1.08 0.260 27.4 -1.62 -0.5477 12 ab
## Semé -0.37 0.333 22.4 -1.06 0.3202 2 b
##
## Results are averaged over the levels of: Sp_Pollinisateurs
## Degrees-of-freedom method: kenward-roger
## Results are given on the log (not the response) scale.
## Confidence level used: 0.95
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM5_G5_mJ <- Interactions_nvis_fl_h %>%
filter(Periode == "Mi-juillet")
LM5_G5_mJ <- lmer(log(nvis_fl_h) ~ Area_gis_m_sq + Temperature + Gestion_moment_5 + Sp_Pollinisateurs + (1|Parcelle), data = CM5_G5_mJ)
eLM5_G5_mJ<-emmeans(LM5_G5_mJ,"Gestion_moment_5")
mcLM5_G5_mJ<-cld(eLM5_G5_mJ,ajust="tukey")
mcLM5_G5_mJ$.group<-as.numeric(mcLM5_G5_mJ$.group)
mcLM5_G5_mJ$group[mcLM5_G5_mJ$.group == 1] <- "a"
mcLM5_G5_mJ$group[mcLM5_G5_mJ$.group == 2] <- "b"
mcLM5_G5_mJ$group[mcLM5_G5_mJ$.group == 12] <- "ab"
mcLM5_G5_mJ## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Tonte tardive -1.852 0.237 41.3 -2.331 -1.373 1 a
## Graminées -1.374 0.570 41.5 -2.525 -0.223 12 ab
## Tonte récente -1.027 0.374 77.9 -1.773 -0.282 12 ab
## Fleuri -0.656 0.234 36.6 -1.129 -0.182 2 b
## Semé -0.424 0.268 25.5 -0.976 0.127 2 b
##
## Results are averaged over the levels of: Sp_Pollinisateurs
## Degrees-of-freedom method: kenward-roger
## Results are given on the log (not the response) scale.
## Confidence level used: 0.95
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.CM5_G5_fJ <- Interactions_nvis_fl_h %>%
filter(Periode == "Fin juillet")
LM5_G5_fJ <- lmer(log(nvis_fl_h) ~ Area_gis_m_sq + Temperature + Gestion_moment_5 + Sp_Pollinisateurs + (1|Parcelle), data = CM5_G5_fJ)
eLM5_G5_fJ<-emmeans(LM5_G5_fJ,"Gestion_moment_5")
mcLM5_G5_fJ<-cld(eLM5_G5_fJ,ajust="tukey")
mcLM5_G5_fJ$.group<-as.numeric(mcLM5_G5_fJ$.group)
mcLM5_G5_fJ$group[mcLM5_G5_fJ$.group == 1] <- "a"
mcLM5_G5_fJ$group[mcLM5_G5_fJ$.group == 2] <- "b"
mcLM5_G5_fJ$group[mcLM5_G5_fJ$.group == 12] <- "ab"
mcLM5_G5_fJ$group[mcLM5_G5_fJ$.group == 23] <- "bc"
mcLM5_G5_fJ$group[mcLM5_G5_fJ$.group == 3] <- "c"
mcLM5_G5_fJ$group[mcLM5_G5_fJ$.group == 123] <- "abc"
mcLM5_G5_fJ## Gestion_moment_5 emmean SE df lower.CL upper.CL .group group
## Tonte récente -1.904 0.441 38.8 -2.797 -1.011 12 ab
## Tonte tardive -1.834 0.229 44.6 -2.295 -1.374 1 a
## Graminées -0.703 0.563 31.0 -1.852 0.446 123 abc
## Fleuri -0.457 0.302 29.3 -1.075 0.161 23 bc
## Semé -0.179 0.313 16.9 -0.841 0.482 3 c
##
## Results are averaged over the levels of: Sp_Pollinisateurs
## Degrees-of-freedom method: kenward-roger
## Results are given on the log (not the response) scale.
## Confidence level used: 0.95
## P value adjustment: tukey method for comparing a family of 5 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.eLM5_nvis_spP<-emmeans(LM5_nvis,"Sp_Pollinisateurs")
mcLM5_nvis_spP<-cld(eLM5_nvis_spP,ajust="tukey")
mcLM5_nvis_spP$.group<-as.numeric(mcLM5_nvis_spP$.group)
mcLM5_nvis_spP$group[mcLM5_nvis_spP$.group == 1] <- "a"
mcLM5_nvis_spP$group[mcLM5_nvis_spP$.group == 2] <- "b"
mcLM5_nvis_spP$group[mcLM5_nvis_spP$.group == 12] <- "ab"
mcLM5_nvis_spP## Sp_Pollinisateurs emmean SE df lower.CL upper.CL .group group
## Apis mellifera -1.38 0.120 74.3 -1.62 -1.142 1 a
## Bombus lapidarius -1.21 0.150 165.0 -1.51 -0.918 12 ab
## Bombus pascuorum -1.20 0.153 171.1 -1.50 -0.893 12 ab
## Hylaeus spec. -1.08 0.167 221.2 -1.41 -0.751 12 ab
## Episyrphus balteatus -1.02 0.176 273.0 -1.37 -0.674 12 ab
## Lasioglossum spec. -0.95 0.137 128.3 -1.22 -0.679 2 b
##
## Results are averaged over the levels of: Periode, Gestion_moment_5
## Degrees-of-freedom method: kenward-roger
## Results are given on the log (not the response) scale.
## Confidence level used: 0.95
## P value adjustment: tukey method for comparing a family of 6 estimates
## significance level used: alpha = 0.05
## NOTE: If two or more means share the same grouping symbol,
## then we cannot show them to be different.
## But we also did not show them to be the same.3.2.6 LMM 6. Tonte
3.2.6.1 S
LM6_Tonte_S <- lmer(S ~ Jour_af_tonte + (1|Site), data = expe_Tonte)
# check_model(LM6_Tonte_S)
# shapiro.test(residuals(LM6_Tonte_S))
LM6_Tonte_S <- lmer((S)^2 ~ Jour_af_tonte + (1|Site), data = expe_Tonte)
# check_model(LM6_Tonte_S)
# shapiro.test(residuals(LM6_Tonte_S))
step(LM6_Tonte_S, direction = "backward")
# (S)^2 ~ Jour_af_tontecheck_model(LM6_Tonte_S)
Anova(LM6_Tonte_S)## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: (S)^2
## Chisq Df Pr(>Chisq)
## Jour_af_tonte 23.293 1 1.391e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1#summary(LM6_Tonte_S)
#LM6_Tonte_S3.2.6.2 Ab
LM6_Tonte_Ab <- lmer(Ab ~ Jour_af_tonte + (1|Site), data = expe_Tonte)
# check_model(LM6_Tonte_Ab)
# shapiro.test(residuals(LM6_Tonte_Ab))
LM6_Tonte_Ab <- lmer(sqrt(Ab) ~ Jour_af_tonte + (1|Site), data = expe_Tonte)
step(LM6_Tonte_Ab, direction = "backward")
# sqrt(Ab) ~ Jour_af_tonte + (1 | Site)check_model(LM6_Tonte_Ab)
Anova(LM6_Tonte_Ab)## Analysis of Deviance Table (Type II Wald chisquare tests)
##
## Response: sqrt(Ab)
## Chisq Df Pr(>Chisq)
## Jour_af_tonte 114.88 1 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1#summary(LM6_Tonte_Ab)
#LM6_Tonte_Ab4 Autres analyses
4.1 Interactions vs pas d’interactions: tests d’homogénéité
Interactions_Gestion$Inter_YN <- ifelse(Interactions_Gestion$N_Interactions == 0, 0, 1)
Interactions_Gestion %>%
group_by(Gestion_2, Inter_YN) %>%
mutate(Inter_YN = as.factor(Inter_YN)) %>%
summarize(n = n()) %>%
ggplot(aes(x = Inter_YN, y = n, color = Gestion_2)) +
geom_point(size = 3)
Interactions_Gestion %>%
group_by(Gestion_moment_5, Inter_YN) %>%
mutate(Inter_YN = as.factor(Inter_YN)) %>%
summarize(n = n()) %>%
ggplot(aes(x = Inter_YN, y = n, color = Gestion_moment_5)) +
geom_jitter(size = 3, width = 0.15) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) 
a <- Interactions_Gestion %>%
group_by(Gestion_moment_5) %>%
summarize(a = n())
b <- Interactions_Gestion %>%
group_by(Gestion_moment_5, Inter_YN) %>%
mutate(Inter_YN = as.factor(Inter_YN)) %>%
summarize(n = n())
c <- full_join(a,b)
c %>%
mutate(prop = n/a) %>%
filter(Inter_YN == 1) %>%
ggplot(aes(x = Gestion_moment_5, y = prop, color = Gestion_moment_5)) +
geom_point(size = 3) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) 
Interactions_Gestion <- Interactions_Gestion %>%
mutate(Inter_YN = as.factor(Inter_YN))
levels(Interactions_Gestion$Inter_YN) <- c("Pas d'interactions","Interactions")
Interactions_Gestion %>%
group_by(Gestion_moment_5, Inter_YN) %>%
mutate(Inter_YN = as.factor(Inter_YN)) %>%
summarize(n = n()) %>%
ggplot(aes(x = Gestion_moment_5, y = n, color = Inter_YN)) +
geom_jitter(size = 3, width = 0.15) 
Interactions_Gestion %>%
group_by(Gestion_moment_5, Inter_YN) %>%
mutate(Inter_YN = as.factor(Inter_YN)) %>%
summarize(n = n()) %>%
ggplot(aes(x = Gestion_moment_5, y = n, color = Gestion_moment_5)) +
facet_wrap(~Inter_YN) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) 
4.2 Espèces de plantes ayant zero interactions
Totalité des quadrats
Pl_zero <- Interactions_Gestion %>%
replace(is.na(.), 0) %>%
group_by(Sp_Plantes) %>%
summarise(data = sum(N_Interactions)) %>%
filter(data == 0)
Pl_zero ## # A tibble: 17 × 2
## Sp_Plantes data
## <chr> <int>
## 1 "" 0
## 2 "Cerastium fontanum" 0
## 3 "Galium mollugo" 0
## 4 "Geranium dissectum" 0
## 5 "Geranium molle" 0
## 6 "Lapsana communis" 0
## 7 "Lathyrus latifolius" 0
## 8 "Plantago lanceolata" 0
## 9 "Ranunculus acris" 0
## 10 "Rhinanthus minor" 0
## 11 "Tanacetum vulgare" 0
## 12 "Trifolium campestre" 0
## 13 "Trifolium dubium" 0
## 14 "Veronica arvensis" 0
## 15 "Veronica persica" 0
## 16 "Veronica serpyllifolia" 0
## 17 "Vicia sativa" 016 espèces de plantes non-visitées, dans aucun quadrat!
4.3 Nombre de changements de fleurs/sp
pour sp de poll. x individus : +ieurs fleurs en séquences
Interactions_indiv <- Interactions_Classes %>%
filter(!is.na(individu)) %>%
mutate(Sp_pollinisateurs = Sp_Pollinisateurs) %>%
unite('Sp_unique', Sp_Pollinisateurs:individu)
seq <- Interactions_indiv %>%
count(Sp_unique) %>%
arrange(desc(n))
indiv_seq <- full_join(Interactions_indiv,seq)
sum <- Interactions_indiv %>%
group_by(Sp_unique) %>%
summarize(sum = sum(N_Interactions)) %>%
arrange(desc(sum))
join_sum <- full_join(Interactions_indiv, sum)
join_sum_n <- full_join(join_sum, indiv_seq)
join_sum_n %>%
ggplot(aes(x= Classe_Poll, y= sum, color = Gestion_2))+
geom_boxplot(alpha = 0.70)
join_sum_n %>%
ggplot(aes(x= Sp_pollinisateurs, y= sum, color = Gestion_2))+
geom_point()
join_sum_n %>%
mutate(n = as.factor(n)) %>%
ggplot(aes(x= Sp_pollinisateurs, y= sum, color = n))+
geom_point()
join_sum_n %>%
mutate(n = as.factor(n)) %>%
ggplot(aes(x= Classe_Poll, y= sum, color = n))+
geom_boxplot(alpha = 0.70)
join_sum_n %>% ggplot(aes(x= Sp_pollinisateurs, y= n, color = Classe_Poll))+
geom_point()
join_sum_n %>% ggplot(aes(x= Classe_Poll, y= n, color = Gestion_2))+
geom_point(size = 4)
Interactions_Gestion %>%
select(Site_gestion_date_Quadrat, Gestion_moment_5, Sp_Plantes, Sp_Pollinisateurs, individu, N_Interactions) %>%
filter(individu == "1315"|
individu == "1434"|
individu == "1325"|
individu == "749"|
individu == "536"|
individu == "1202"|
individu == "755")## Site_gestion_date_Quadrat Gestion_moment_5 Sp_Plantes
## 1 CAV_Fauche_finJuillet_Q1 Semé Centaurea jacea
## 2 CAV_Fauche_finJuillet_Q1 Semé Trifolium pratense
## 3 CAV_Fauche_finJuillet_Q1 Semé Knautia arvensis
## 4 CAV_Fauche_finJuillet_Q2 Semé Betonica officinalis
## 5 CAV_Fauche_finJuillet_Q2 Semé Prunella vulgaris
## 6 CAV_Fauche_finJuillet_Q2 Semé Centaurea jacea
## 7 CAV_Fauche_miJuillet_Q2 Semé Centaurea jacea
## 8 CAV_Fauche_miJuillet_Q2 Semé Betonica officinalis
## 9 CAV_Fauche_miJuillet_Q2 Semé Betonica officinalis
## 10 CAV_Fauche_miJuillet_Q2 Semé Prunella vulgaris
## 11 CAV_Fauche_miJuillet_Q2 Semé Prunella vulgaris
## 12 CAV_Fauche_miJuillet_Q2 Semé Trifolium pratense
## 13 Kellner_Fauche_Juin_Q3 Semé Achillea millefolium
## 14 Kellner_Fauche_Juin_Q3 Semé Jacobaea vulgaris
## 15 Kellner_Fauche_Juin_Q3 Semé Origanum vulgare
## 16 Kellner_Fauche_miJuillet_Q1 Semé Hypericum perforatum
## 17 Kellner_Fauche_miJuillet_Q1 Semé Jacobaea vulgaris
## 18 Kellner_Fauche_miJuillet_Q1 Semé Lotus corniculatus
## 19 Reaumur_Fauche_finJuillet_Q1 Semé Cichorium intybus
## 20 Reaumur_Fauche_finJuillet_Q1 Semé Cirsium arvense
## 21 Reaumur_Fauche_finJuillet_Q1 Semé Centaurea jacea
## Sp_Pollinisateurs individu N_Interactions
## 1 Bombus pascuorum 1315 1
## 2 Bombus pascuorum 1315 1
## 3 Bombus pascuorum 1315 2
## 4 Bombus pascuorum 1325 1
## 5 Bombus pascuorum 1325 4
## 6 Bombus pascuorum 1325 1
## 7 Bombus pascuorum 749 2
## 8 Bombus pascuorum 749 1
## 9 Bombus pascuorum 755 9
## 10 Bombus pascuorum 749 3
## 11 Bombus pascuorum 755 3
## 12 Bombus pascuorum 755 3
## 13 Helophilus pendulus 536 1
## 14 Helophilus pendulus 536 1
## 15 Helophilus pendulus 536 1
## 16 Megachile spec. 1202 1
## 17 Megachile spec. 1202 3
## 18 Megachile spec. 1202 1
## 19 Bombus lapidarius 1434 1
## 20 Bombus lapidarius 1434 2
## 21 Bombus lapidarius 1434 2length(unique(Interactions_indiv$Sp_Plantes))## [1] 335 Graphiques finaux
5.1 Heure
ggplot(Inv) +
aes(x = Heure_debut, y = Ab_Poll, color = Gestion_moment_5) +
geom_point()+
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
scale_x_discrete(breaks = c("09:00:00","10:00:00","11:00:00","12:00:00","13:00:00","14:00:00","15:00:00","16:00:00", "17:00:00")) +
labs(x = "Heure",
y = "Abondance en pollinisateurs",
color = "Type de gestion") +
theme(legend.position="bottom",
legend.title = element_text(size= 12),
legend.text = element_text(size=10),
axis.text=element_text(size=12),
axis.title=element_text(size=14))
5.2 Meteo
ggplot(Inv) +
aes(x = Gestion_moment_5, y = Ab_Poll, color = Meteo) +
geom_boxplot(alpha = 0.70) +
scale_color_manual(values = c("Nuageux" = "#7f908c",
"Alternances" = "#79ccbd",
"Soleil" = "#fbcb09")) +
labs(x = "Type de gestion",
y = "Abondance en pollinisateurs",
color = "Météo") +
theme(legend.position="bottom",
legend.title = element_text(size= 12),
legend.text = element_text(size=10),
axis.text=element_text(size=12),
axis.title=element_text(size=14))
5.3 Courbes d’accumulation
5.3.1 Plantes
ggplot(data=accum.long_Plantes_G2, aes(x = Sites, y = Richness, ymax = UPR, ymin = LWR)) +
scale_x_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
scale_y_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
geom_line(aes(colour=Grouping), size=1.5) +
geom_point(data=subset(accum.long_Plantes_G2, labelit==TRUE),
aes(colour=Grouping, shape=Grouping), size=2.5) +
scale_shape_manual(values = c("Fauche" = 16,
"Tonte" = 5)) +
geom_ribbon(aes(colour=Grouping, fill=after_scale(alpha(colour, alpha=0.15))),
show.legend=FALSE) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
labs(x = "Nombre d'échantillonages", y = "Nombre d'espèces de plantes", colour = "Type de gestion", shape = "Type de gestion") +
theme(legend.position = c(0.85,0.07),
legend.text = element_text(size=10),
axis.text=element_text(size=12),
axis.title=element_text(size=14)) +
ggplot(data=accum.long_Plantes_G5, aes(x = Sites, y = Richness, ymax = UPR, ymin = LWR)) +
scale_x_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
scale_y_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
geom_line(aes(colour=Grouping), size=1.5) +
geom_point(data=subset(accum.long_Plantes_G5, labelit==T),
aes(colour=Grouping, shape=Grouping), size=2.5) +
scale_shape_manual(values = c("Graminées" = 16,
"Fleuri" = 17,
"Semé" = 15,
"Tonte récente" = 5,
"Tonte tardive" = 6)) +
geom_ribbon(aes(colour=Grouping, fill=after_scale(alpha(colour, alpha=0.15))),
show.legend=FALSE) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Nombre d'échantillonages", y = "Nombre d'espèces de plantes", colour = "Type de gestion", shape = "Type de gestion") +
theme(legend.position = c(0.82,0.12),
legend.text = element_text(size=10),
axis.text=element_text(size=12),
axis.title=element_text(size=14)) 
5.3.2 Pollinisateurs
ggplot(data=accum.long_Poll_G2, aes(x = Sites, y = Richness, ymax = UPR, ymin = LWR)) +
scale_x_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
scale_y_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
geom_line(aes(colour=Grouping), size=1.5) +
geom_point(data=subset(accum.long_Poll_G2, labelit==TRUE),
aes(colour=Grouping, shape=Grouping), size=2.5) +
scale_shape_manual(values = c("Fauche" = 16,
"Tonte" = 5)) +
geom_ribbon(aes(colour=Grouping, fill=after_scale(alpha(colour, alpha=0.15))),
show.legend=FALSE) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
labs(x = "Nombre d'échantillonages", y = "Nombre d'espèces de pollinisateurs", colour = "Type de gestion", shape = "Type de gestion") +
theme(legend.position = c(0.85,0.07),
legend.text = element_text(size=10),
axis.text=element_text(size=12),
axis.title=element_text(size=14)) +
ggplot(data=accum.long_Poll_G5, aes(x = Sites, y = Richness, ymax = UPR, ymin = LWR)) +
scale_x_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
scale_y_continuous(sec.axis = dup_axis(labels=NULL, name=NULL)) +
geom_line(aes(colour=Grouping), size=1.5) +
geom_point(data=subset(accum.long_Poll_G5, labelit==T),
aes(colour=Grouping, shape=Grouping), size=2.5) +
scale_shape_manual(values = c("Graminées" = 16,
"Fleuri" = 17,
"Semé" = 15,
"Tonte récente" = 5,
"Tonte tardive" = 6)) +
geom_ribbon(aes(colour=Grouping, fill=after_scale(alpha(colour, alpha=0.15))),
show.legend=FALSE) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Nombre d'échantillonages", y = "Nombre d'espèces de pollinisateurs", colour = "Type de gestion", shape = "Type de gestion") +
theme(legend.position = c(0.82,0.12),
legend.text = element_text(size=10),
axis.text=element_text(size=12),
axis.title=element_text(size=14)) 
5.4 Analyses multivariées
Voir précédement
5.5 Modèles Inventaires
5.5.1 LM1_S_Pl
Inv %>% ggplot(aes (x = Gestion_2, y = S_Plant)) +
geom_boxplot(aes (color = Gestion_2), alpha = 0.70) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
labs(#title = "Richesse spécifique en plantes en fonction des types de gestion",
x = "Type de gestion", y = "Richesse spécifique en plantes") +
theme(legend.position = "none",
axis.text=element_text(size=12),
axis.title=element_text(size=14)) +
geom_text(aes(label = "b", y = 14, x = 1),color="black") +
geom_text(aes(label = "a", y = 7, x = 2),color="black")
5.5.2 LM1_Ab_Pl
Inv %>% ggplot(aes (x = Periode, y = Ab_Plant)) +
geom_boxplot(aes (color = Periode), alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(#title = "Abondance en plantes en fonction des types de gestion",
x = "Période", y = "Abondance en plantes") +
theme(legend.position = "none",
axis.text=element_text(size=12),
axis.title=element_text(size=14)) +
geom_text(aes(label = "b", y = 265, x =1),color="black") +
geom_text(aes(label = "a", y = 210, x =2),color="black") +
geom_text(aes(label = "a", y = 160, x =3),color="black") +
Inv %>% ggplot(aes (x = Temperature, y = Ab_Plant)) +
geom_point() +
geom_smooth(method = "lm", se = T) +
labs(#title = "Richesse spécifique en pollinisateurs en fonction des types de gestion",
x = "Température (°C)", y = "Abondance en plantes") +
theme(axis.text=element_text(size=12),
axis.title=element_text(size=14))
5.5.3 LM1_S_Po
Inv %>% ggplot(aes (x = Gestion_2, y = S_Poll)) +
geom_boxplot(aes (color = Gestion_2), alpha = 0.70) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
labs(#title = "Richesse spécifique en pollinisateurs en fonction des types de gestion",
x = "Type de gestion", y = "Richesse spécifique en pollinisateurs") +
theme(legend.position = "none",
axis.text=element_text(size=12),
axis.title=element_text(size=14)) +
geom_text(aes(label = "b", y = 21.5, x =1),color="black") +
geom_text(aes(label = "a", y = 10, x =2),color="black")
5.5.4 LM1_Ab_Po
Inv %>% ggplot(aes (x = Gestion_2, y = Ab_Poll)) +
geom_boxplot(aes (color = Gestion_2), alpha = 0.70) +
scale_color_manual(values = c("Fauche" = "#aa1e0f",
"Tonte" = "#12661f")) +
labs(#title = "Abondance en pollinisateurs en fonction des types de gestion",
x = "Type de gestion", y = "Abondance en pollinisateurs") +
theme(legend.position = "none",
axis.text=element_text(size=12),
axis.title=element_text(size=14)) +
geom_text(aes(label = "b", y = 55, x =1),color="black") +
geom_text(aes(label = "a", y = 19, x =2),color="black")
5.5.5 LM2_S_Pl
Inv %>% ggplot(aes (x = Periode, y = S_Plant)) +
geom_boxplot(aes (color = Periode), alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(#title = "Abondance de plantes en fonction des types de gestion",
x = "Période", y = "Richesse spécifique en plantes", color = "Période") +
theme(legend.position = "none",
axis.text=element_text(size=12),
axis.title=element_text(size=14)) +
geom_text(aes(label = "a", y = 11.5, x =1),color="black") +
geom_text(aes(label = "a", y = 9.5, x =2),color="black") +
geom_text(aes(label = "a", y = 7.5, x =3),color="black") +
Inv %>% ggplot(aes (x = Gestion_moment_5, y = S_Plant)) +
geom_boxplot(aes (color = Gestion_moment_5), alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",#6dcf20
"Tonte tardive" = "#2b790c")) + #12661f
labs(#title = "Richesse spécifique en plantes en fonction des types de gestion",
x = "Type de gestion", y = "Richesse spécifique en plantes") +
theme(legend.position = "none",
axis.text=element_text(size=10),
axis.title=element_text(size=14)) +
geom_text(aes(label = "a", y = 6, x =1),color="black") +
geom_text(aes(label = "a", y = 5, x =4),color="black") +
geom_text(aes(label = "b", y = 7, x =5),color="black") +
geom_text(aes(label = "bc", y = 12, x =2),color="black") +
geom_text(aes(label = "c", y = 18, x =3),color="black") 
5.5.6 LM2_Ab_Pl
Inv %>% ggplot(aes (x = Periode, y = Ab_Plant)) +
geom_boxplot(aes (color = Periode), alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(#title = "Abondance de plantes en fonction des types de gestion",
x = "Période", y = "Abondance en plantes", color = "Période") +
theme(legend.position = "none",
axis.text=element_text(size=12),
axis.title=element_text(size=14)) +
geom_text(aes(label = "b", y = 265, x =1),color="black") +
geom_text(aes(label = "a", y = 210, x =2),color="black") +
geom_text(aes(label = "a", y = 160, x =3),color="black") +
Inv %>% ggplot(aes (x = Gestion_moment_5, y = Ab_Plant)) +
geom_boxplot(aes (color = Gestion_moment_5), alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(#title = "Abondance de plantes en fonction des types de gestion",
x = "Type de gestion", y = "Abondance en plantes", color = "Type de gestion") +
theme(legend.position = "none",
axis.text=element_text(size=10),
axis.title=element_text(size=14)) +
geom_text(aes(label = "a", y = 60, x =1),color="black") +
geom_text(aes(label = "b", y = 270, x =2),color="black") +
geom_text(aes(label = "b", y = 270, x =3),color="black") +
geom_text(aes(label = "a", y = 110, x =4),color="black") +
geom_text(aes(label = "b", y = 210, x =5),color="black") 
5.5.7 LM2_S_Po
Inv %>% ggplot(aes (x = Periode, y = S_Poll)) +
geom_boxplot(aes (color = Periode), alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(x = "Période", y = "Richesse spécifique en pollinisateurs", color = "Type de gestion") +
theme(legend.position = "none",
axis.text=element_text(size=10),
axis.title=element_text(size=14)) +
geom_text(aes(label = "a", y = 16.5, x = 1)) +
geom_text(aes(label = "a", y = 19.5, x = 2)) +
geom_text(aes(label = "a", y = 16.5, x = 3)) +
Inv %>% ggplot(aes (x = Gestion_moment_5, y = S_Poll)) +
geom_boxplot(aes (color = Gestion_moment_5), alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Type de gestion", y = "Richesse spécifique en pollinisateurs", color = "Type de gestion") +
theme(legend.position = "none",
axis.text=element_text(size=10),
axis.title=element_text(size=14)) +
geom_text(aes(label = "ab", y = 12.5, x = 1)) +
geom_text(aes(label = "c", y = 17.5, x = 2)) +
geom_text(aes(label = "c", y = 21.5, x = 3)) +
geom_text(aes(label = "a", y = 5.5, x = 4)) +
geom_text(aes(label = "b", y = 9.5, x = 5))
data_text <- data.frame(label = c("a", "b", "b","a", "a",
"ab", "cd", "d","a", "bc",
"b", "b", "b","a", "a"),
Periode = rep(c("Juin", "Mi-juillet", "Fin juillet"), each = 5),
x = c(1,2,3,4,5,
1,2,3,4,5,
1,2,3,4,5),
y = c(9,14,17,6,10,
10,18,20,5,6,
13,10,11,3,6))
data_text$Periode <- fct_relevel(data_text$Periode, c("Juin", "Mi-juillet", "Fin juillet"))
Inv %>% ggplot(aes (x = Gestion_moment_5, y = S_Poll)) +
facet_wrap(~Periode)+
geom_boxplot(aes (color = Gestion_moment_5), alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Type de gestion", y = "Richesse spécifique en pollinisateurs") +
theme(legend.position = "none",
#axis.title.x=element_blank(),
legend.title = element_text(size=11),
legend.text = element_text(size=10),
axis.text.y = element_text(size=10),
axis.text.x = element_text(size=8),
axis.title = element_text(size=14),
strip.text = element_text(size=10)) +
geom_text(data = data_text %>% filter(Periode == Periode),
mapping = aes(x = x, y = y, label = label),
size = 4, fontface = "bold")
5.5.8 LM2_Ab_Po
Inv %>% ggplot(aes (x = Periode, y = Ab_Poll)) +
geom_boxplot(aes (color = Periode), alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(x = "Période", y = "Abondance en pollinisateurs", color = "Type de gestion") +
theme(legend.position = "none",
axis.text=element_text(size=10),
axis.title=element_text(size=14)) +
geom_text(aes(label = "a", y = 42.5, x = 1)) +
geom_text(aes(label = "a", y = 45, x = 2)) +
geom_text(aes(label = "a", y = 29, x = 3)) +
Inv %>% ggplot(aes (x = Gestion_moment_5, y = Ab_Poll)) +
geom_boxplot(aes (color = Gestion_moment_5), alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Type de gestion", y = "Abondance en pollinisateurs", color = "Type de gestion") +
theme(legend.position = "none",
axis.text=element_text(size=10),
axis.title=element_text(size=14)) +
geom_text(aes(label = "ab", y = 17.5, x = 1)) +
geom_text(aes(label = "c", y = 45, x = 2)) +
geom_text(aes(label = "d", y = 62, x = 3)) +
geom_text(aes(label = "a", y = 9, x = 4)) +
geom_text(aes(label = "b", y = 19, x = 5))
data_text <- data.frame(label = c("ab", "c", "c","a", "b",
"ab", "c", "d","a", "bc",
"b", "ab", "b","a", "a"),
Periode = rep(c("Juin", "Mi-juillet", "Fin juillet"), each = 5),
x = c(1,2,3,4,5,
1,2,3,4,5,
1,2,3,4,5),
y = c(16,40,43,9,29,
19,45,63,6,20,
45,25,40,6,19))
data_text$Periode <- fct_relevel(data_text$Periode, c("Juin", "Mi-juillet", "Fin juillet"))
Inv %>% ggplot(aes (x = Gestion_moment_5, y = Ab_Poll)) +
facet_wrap(~Periode)+
geom_boxplot(aes (color = Gestion_moment_5), alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Type de gestion", y = "Abondance en pollinisateurs", color = "Type de gestion") +
theme(legend.position = "none",
legend.title = element_text(size=11),
legend.text = element_text(size=10),
axis.text.y = element_text(size=10),
axis.text.x = element_text(size=8),
axis.title = element_text(size=14),
strip.text = element_text(size=10)) +
geom_text(data = data_text %>% filter(Periode == Periode),
mapping = aes(x = x, y = y, label = label),
size = 4, fontface = "bold")
5.6 Modèles Interactions
5.6.1 LM3_Int : N_Int & qtté fleurs
Interactions_Gestion %>% ggplot(aes (x = Qtte_Plantes, y = N_Interactions)) +
geom_point() +
geom_smooth(method = "lm") +
theme(legend.position = "none",
axis.text=element_text(size=10),
axis.title=element_text(size=14)) +
labs(x = "Quantité d'unités florales", y = "Nombre d'interactions") +
Interactions_Gestion %>% ggplot(aes (x = Temperature, y = N_Interactions)) +
geom_point() +
geom_smooth(method = "lm") +
theme(legend.position = "none",
axis.text=element_text(size=10),
axis.title=element_text(size=14)) +
labs(x = "Température (°C)", y = "Nombre d'interactions") +
Interactions_Gestion %>% ggplot(aes (x = Periode, y = N_Interactions)) +
geom_boxplot(aes (color = Periode), alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
theme(legend.position = "none",
axis.text=element_text(size=10),
axis.title=element_text(size=14)) +
geom_text(aes(label = "a", y = 25, x = 1)) +
geom_text(aes(label = "b", y = 26.5, x = 2)) +
geom_text(aes(label = "b", y = 28, x = 3)) +
labs(x = "Période", y = "Nombre d'interactions")
Interactions_Gestion %>% ggplot(aes (x = Qtte_Plantes, y = N_Interactions)) +
geom_point(aes (color = Gestion_moment_5)) +
geom_smooth(aes (color = Gestion_moment_5), method = "lm", se = T) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
theme(legend.position = "bottom",
axis.text=element_text(size=10),
axis.title=element_text(size=14)) +
labs(x = "Quantité d'unités florales", y = "Nombre d'interactions", color = "Type de gestion")
data_text <- data.frame(label = c("a", "b", "ab",
"a", "b", "ab",
"a", "a", "a",
"a", "a", "a",
"a", "b", "ab"),
Gestion_moment_5 = rep(c("Graminées", "Fleuri", "Semé", "Tonte récente", "Tonte tardive"), each = 3),
x = c(1,2,3,
1,2,3,
1,2,3,
1,2,3,
1,2,3),
y = c(8,8,8,
10,18,7,
11,13,14,
4,11,9,
25,27,28))
data_text$Gestion_moment_5 <- fct_relevel(data_text$Gestion_moment_5, c("Graminées", "Fleuri", "Semé", "Tonte récente", "Tonte tardive"))
Interactions_Gestion %>% ggplot(aes (x = Periode, y = N_Interactions)) +
facet_wrap(~Gestion_moment_5) +
geom_boxplot(aes (color = Periode), alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(x = "Période", y = "Nombre d'interactions", color = "Type de gestion") +
theme(legend.position = "none",
legend.title = element_text(size=11),
legend.text = element_text(size=10),
axis.text.y = element_text(size=10),
axis.text.x = element_text(size=8),
axis.title = element_text(size=14),
strip.text = element_text(size=10)) +
geom_text(data = data_text %>% filter(Gestion_moment_5 == Gestion_moment_5),
mapping = aes(x = x, y = y, label = label),
size = 4, fontface = "bold")
5.6.2 LM4_Int : sum(N_Int)
Température: voir LM3
Interactions_Gestion %>%
group_by(Site_gestion_date, Periode, Gestion_moment_5, Temperature) %>%
summarize(n= sum(N_Interactions)) %>%
ggplot(aes (x = Temperature, y = n)) +
geom_point() +
geom_smooth(method = "lm") +
labs(x = "Température (°C)", y="Somme des interactions") +
theme(axis.text=element_text(size=10),
axis.title=element_text(size=14))
Interactions_Gestion %>%
group_by(Site_gestion_date, Periode, Gestion_moment_5) %>%
summarize(n= sum(N_Interactions)) %>%
ggplot(aes (x = Periode, y = n)) +
geom_boxplot(aes (color = Periode), alpha = 0.70) +
scale_color_manual(values = c(Juin = "#74a9cf",
`Mi-juillet` = "#2b8cbe",
`Fin juillet` = "#045a8d")) +
labs(x = "Période", y = "Somme des interactions", color = "Type de gestion") +
theme(legend.position = "none",
axis.text=element_text(size=10),
axis.title=element_text(size=14)) +
geom_text(aes(label = "a", y = 137, x = 1)) +
geom_text(aes(label = "a", y = 155, x = 2)) +
geom_text(aes(label = "a", y = 75, x = 3)) +
Interactions_Gestion %>%
group_by(Site_gestion_date, Periode, Gestion_moment_5) %>%
summarize(n= sum(N_Interactions)) %>%
ggplot(aes (x = Gestion_moment_5, y = n)) +
geom_boxplot(aes (color = Gestion_moment_5), alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Type de gestion", y = "Somme des interactions", color = "Type de gestion") +
theme(legend.position = "none",
axis.text=element_text(size=10),
axis.title=element_text(size=14)) +
geom_text(aes(label = "ab", y = 27, x = 1)) +
geom_text(aes(label = "b", y = 91, x = 2)) +
geom_text(aes(label = "c", y = 198, x = 3)) +
geom_text(aes(label = "a", y = 48, x = 4)) +
geom_text(aes(label = "b", y = 140, x = 5))
data_text <- data.frame(label = c("a", "b", "b","a", "b",
"a", "b", "c","a", "b",
"ab", "ab", "b","a", "a"),
Periode = rep(c("Juin", "Mi-juillet", "Fin juillet"), each = 5),
x = c(1,2,3,4,5,
1,2,3,4,5,
1,2,3,4,5),
y = c(15,93,85,28,123,
25,80,198,10,113,
98,43,125,19,50))
data_text$Periode <- fct_relevel(data_text$Periode, c("Juin", "Mi-juillet", "Fin juillet"))
Interactions_Gestion %>%
group_by(Site_gestion_date, Periode, Gestion_moment_5) %>%
summarize(n= sum(N_Interactions)) %>%
ggplot(aes (x = Gestion_moment_5, y = n)) +
facet_wrap(~Periode)+
geom_boxplot(aes (color = Gestion_moment_5), alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Type de gestion", y = "Somme des interactions", color = "Type de gestion") +
theme(legend.position = "none",
legend.title = element_text(size=11),
legend.text = element_text(size=10),
axis.text.y = element_text(size=10),
axis.text.x = element_text(size=8),
axis.title = element_text(size=14),
strip.text = element_text(size=10)) +
geom_text(data = data_text %>% filter(Periode == Periode),
mapping = aes(x = x, y = y, label = label),
size = 4, fontface = "bold")
5.6.3 LM5_nvis : nvis_fl_h
#voir mod précédent effet t°, Periode
Interactions_nvis_fl_h %>%
ggplot(aes (x = Periode, y = nvis_fl_h)) +
facet_wrap(~Sp_Pollinisateurs) +
geom_boxplot(aes (color = Gestion_moment_5), alpha = 0.70) +
scale_y_continuous(trans='log10',
breaks=c(0.010, 0.100,1.000, 10),
labels=c("0.01","0.1","1", "10")) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Période", y = "Nombre de visites par unité florale par heure (nvis/uf.h)", color= "Type de gestion") +
theme(legend.position = "bottom",
legend.title = element_text(size=11),
legend.text = element_text(size=10),
axis.text = element_text(size=11),
axis.title = element_text(size=14),
strip.text = element_text(size=10, face = "italic"))
5.7 Interactions ou non
Interactions_Gestion %>%
group_by(Gestion_moment_5) %>%
summarize(tot = n())## # A tibble: 5 × 2
## Gestion_moment_5 tot
## <fct> <int>
## 1 Graminées 108
## 2 Fleuri 652
## 3 Semé 765
## 4 Tonte récente 78
## 5 Tonte tardive 465E_O <- data.frame(Gestion_moment_5 = c(rep(c("Graminées","Graminées","Fleuri", "Fleuri","Semé","Semé","Tonte récente","Tonte récente","Tonte tardive","Tonte tardive"),2)),
EO = c(rep("O",10), rep("E",10)),
Inter_YN= c(rep(c("Pas d'interactions","Interactions"),10)),
EO_YN = c(rep(c("Pas d'interactions, observé","Interactions, observé"),5), rep(c("Pas d'interactions, attendu","Interactions, attendu"),5)),
totEO = c(16,92,89,563,99,666,27,51,87,378,16.61,91.39,100.26,551.74,117.64,647.36,11.99,66.01,71.50,393.50))
# n <- Interactions_Gestion %>%
# group_by(Gestion_moment_5, Inter_YN) %>%
# mutate(Inter_YN = as.factor(Inter_YN)) %>%
# summarize(n = n())
# YN <- full_join(tot,n)
# YN %>%
# ggplot(aes(x = Gestion_moment_5, y= n, color = Gestion_moment_5, shape = Inter_YN)) +
# geom_point(size = 4) +
# scale_color_manual(guide = F, values = c("Graminées" = "#fbcb09",
# "Fleuri" = "#ff7207",
# "Semé" = "#de1e21",
# "Tonte récente" = "#6abe1d",
# "Tonte tardive" = "#2b790c")) +
# scale_shape_manual(values = c("Pas d'interactions" = 15,
# "Interactions" = 17)) +
# guides(shape = guide_legend(title = "")) +
# labs(x = "Type de gestion", y = "Compte") +
# theme(legend.position = "bottom",
# axis.text=element_text(size=10),
# axis.title=element_text(size=14),
# strip.text = element_text(size = 12))
E_O$Gestion_moment_5 <- fct_relevel(E_O$Gestion_moment_5,c("Graminées", "Fleuri", "Semé", "Tonte récente", "Tonte tardive"))
E_O %>%
ggplot(aes(x = Gestion_moment_5, y = totEO, color = Gestion_moment_5, shape = EO_YN)) +
geom_point(size = 4) +
scale_color_manual(guide = F, values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
scale_shape_manual(values = c("Pas d'interactions, observé" = 16,
"Interactions, observé" = 17,
"Pas d'interactions, attendu" = 1,
"Interactions, attendu" = 2)) +
guides(shape = guide_legend(title = "")) +
labs(x = "Type de gestion", y = "Compte") +
theme(legend.position = "bottom",
axis.text=element_text(size=14),
axis.title=element_text(size=16),
legend.text = element_text(size = 13))
E_O %>%
filter(EO == "O") %>%
ggplot()+ aes(x = Gestion_moment_5, y = totEO, fill = Inter_YN) +
geom_col(alpha = 0.5) +
scale_fill_manual(values = c("Interactions" = "#3da834",
"Pas d'interactions" = "#c54126")) +
# scale_color_manual(guide = F, values = c("Graminées" = "#fbcb09",
# "Fleuri" = "#ff7207",
# "Semé" = "#de1e21",
# "Tonte récente" = "#6abe1d",
# "Tonte tardive" = "#2b790c")) +
geom_segment(x = 0.5, xend = 1.5, y= 16.61, yend = 16.61, color="#fbcb09",linewidth=1)+
geom_segment(x = 1.5, xend = 2.5,y= 100.26,yend= 100.26, color = "#ff7207",linewidth=1)+
geom_segment(x = 2.5, xend = 3.5,y= 117.64,yend= 117.64, color = "#de1e21",linewidth=1)+
geom_segment(x = 3.5, xend = 4.5,y= 11.99,yend= 11.99, color = "#6abe1d",linewidth=1)+
geom_segment(x = 4.5, xend = 5.5,y= 71.50,yend= 71.50, color = "#2b790c",linewidth=1) +
theme(legend.position = c(0.9,0.97),
legend.text = element_text(size=10),
axis.text=element_text(size=12),
axis.title=element_text(size=14)) +
labs(x= "Type de gestion", y = "Compte", fill = "")
5.8 Nombre de changements de fleurs/sp
join_sum_n <- join_sum_n %>%
mutate(n = as.factor(n))
levels(join_sum_n$n) <- c("1 espèce visitée", "2 espèces visitées", "3 espèces visitées")
join_sum_n %>%
ggplot(aes(x = Classe_Poll, y = sum))+
facet_wrap(~n) +
geom_boxplot(aes (color = Gestion_moment_5), alpha = 0.70) +
scale_color_manual(values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Catégories de pollinisateurs", y = "Nombre d'unités florales visitées en séquence", color = "Type de gestion", title = "Nombre d'espèces de plantes en fleur différentes visitées") +
theme(legend.position = "bottom",
axis.text.x = element_text(size = 6),
axis.text.y=element_text(size=10),
axis.title=element_text(size=14),
strip.text = element_text(size = 12),
axis.title.y=element_text(angle=-90, vjust=0.5))
5.9 Réseaux Plantes_pollinisateurs
plotweb(network_fauche, high.lablength= 35, low.lablength=27, col.high=c("#aa1e0f"),col.low=c("#aa1e0f"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
plotweb(network_tonte, high.lablength= 35, low.lablength=27, col.high=c("#12661f"),col.low=c("#12661f"),text.rot=90,y.width.high=0.07, y.width.low=0.07,y.lim=c(-1,3.5))
plotweb(network_gram, high.lablength= 35, low.lablength=27, col.high=c("#fbcb09"),col.low=c("#fbcb09"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
plotweb(network_fleuri, high.lablength= 35, low.lablength=27, col.high=c("#ff7207"),col.low=c("#ff7207"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
plotweb(network_seme, high.lablength= 35, low.lablength=27, col.high=c("#de1e21"),col.low=c("#de1e21"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
plotweb(network_TonRec, high.lablength= 35, low.lablength=27, col.high=c("#6abe1d"), col.low=c("#6abe1d"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
plotweb(network_TonTard, high.lablength= 35, low.lablength=27, col.high=c("#2b790c"), col.low=c("#2b790c"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
plotweb(network_Juin, high.lablength= 35, low.lablength=27, col.high=c("#74a9cf"), col.low=c("#74a9cf"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
plotweb(network_miJuil, high.lablength= 35, low.lablength=27, col.high=c("#2b8cbe"), col.low=c("#2b8cbe"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
plotweb(network_finJuil, high.lablength= 35, low.lablength=27, col.high=c("#045a8d"), col.low=c("#045a8d"), text.rot=90, y.width.high=0.07, y.width.low=0.07, y.lim=c(-1,3.5))
Table_NW_P ## Juin Mi-juillet Fin juillet
## connectance 0.1352785 0.1303030 0.1064815
## web asymmetry 0.3809524 0.3924051 0.3846154
## weighted nestedness 0.5246653 0.5805779 0.5296713
## linkage density 4.7004410 4.2604178 3.6859752
## Fisher alpha 61.9488092 44.7255342 40.7488484
## Shannon diversity 3.7803507 3.5367435 3.1842605
## interaction evenness 0.5165444 0.4922134 0.4442924
## H2 0.5157954 0.4665388 0.5867317
## robustness.HL 0.7788051 0.7668089 0.7323744
## robustness.LL 0.6363928 0.6371332 0.5838619
## generality.HL 3.9909709 3.5874657 3.0955332
## vulnerability.LL 5.4099111 4.9333699 4.2764171# %>%
# kbl(caption = "Période") %>%
# kable_classic(full_width = F, html_font = "Cambria") #%>% kable_styling() %>%
#save_kable(file = "Output/Tableau/NW_P.pdf")
Table_NW_G2 ## Fauche Tonte
## connectance 0.1181478 0.2943723
## web asymmetry 0.4590164 0.6500000
## weighted nestedness 0.6686565 0.4727312
## linkage density 6.9668799 2.1539505
## Fisher alpha 102.2384405 13.8299152
## Shannon diversity 4.5609652 1.6289212
## interaction evenness 0.5711813 0.2993010
## H2 0.4291655 0.6866345
## robustness.HL 0.7905750 0.8376929
## robustness.LL 0.6427648 0.5099233
## generality.HL 5.2007056 1.3405415
## vulnerability.LL 8.7330541 2.9673596# %>%
# kbl(caption = "Gestion - biclassification") %>%
# kable_classic(full_width = F, html_font = "Cambria") # %>% kable_styling() %>%
# save_kable(file = "Output/Tableau/NW_G2.pdf")
Table_NW_G5 ## Graminées Fleuri Semé Tonte récente
## connectance 0.2407407 0.1142534 0.1418764 0.4722222
## web asymmetry 0.6363636 0.4468085 0.4250000 0.6000000
## weighted nestedness 0.3958483 0.5646260 0.6423572 0.3757457
## linkage density 4.5620168 5.5351438 4.9818744 1.9228986
## Fisher alpha 15.7290348 71.1656509 54.1504976 4.6059563
## Shannon diversity 3.2435071 4.2667950 3.8437780 1.2878123
## interaction evenness 0.6375323 0.5706099 0.5354536 0.3593709
## H2 0.7336450 0.5114385 0.4496612 0.8104420
## robustness.HL 0.8052689 0.7730934 0.7710849 0.7025725
## robustness.LL 0.4016080 0.6190162 0.6374439 0.3201852
## generality.HL 1.4745200 3.8796971 3.3251203 1.1012144
## vulnerability.LL 7.6495137 7.1905906 6.6386285 2.7445828
## Tonte tardive
## connectance 0.2857143
## web asymmetry 0.6500000
## weighted nestedness 0.4584951
## linkage density 2.1270847
## Fisher alpha 13.6693767
## Shannon diversity 1.6404026
## interaction evenness 0.3014106
## H2 0.6990216
## robustness.HL 0.8498949
## robustness.LL 0.5047716
## generality.HL 1.3308515
## vulnerability.LL 2.9233178# %>%
# kbl(caption = "Gestion - pentaclassification") %>%
# kable_classic(full_width = F, html_font = "Cambria") #%>% kable_styling() %>%
# save_kable(file = "Output/Tableau/NW_G5.pdf")
# write.csv2(Table_NW_G5, file = "Output/Tableau/NW_G5.csv")
# write.csv2(Table_NW_G2, file = "Output/Tableau/NW_G2.csv")
# write.csv2(Table_NW_P, file = "Output/Tableau/NW_P.csv")5.10 Tonte
5.10.1 LM6_Tonte_S
Tonte_pl %>% ggplot(aes (x = Jour_af_tonte, y = S, color = Site)) +
geom_point() +
geom_smooth(se = F)+
scale_x_continuous(n.breaks=17)+
labs(x = "Jour après la tonte", y = "Richesse spécifique en plantes") +
theme(legend.position = "bottom",
axis.text=element_text(size=10),
axis.title=element_text(size=14))
5.10.2 LM6_Tonte_Ab
Tonte_pl %>% ggplot(aes (x = Jour_af_tonte, y = Qtté, color=Espèces)) +
facet_wrap(~Site) +
geom_smooth(se = F)+
scale_color_viridis_d(option = "viridis", direction = -1) +
geom_point(shape = "circle", size = 1.5) +
scale_x_continuous(n.breaks=17) +
labs(x = "Jour après la tonte", y = "Abondance en plantes") +
theme(legend.position = "right",
axis.text=element_text(size=10),
axis.title=element_text(size=14),
strip.text = element_text(size = 12)) +
theme_classic()
5.11 Estimation % diversité totale
Estim <- Inventaire %>%
rownames_to_column(var = "temp")
Estim <- Estim[,c(12, 20:167)]
Estim <- aggregate(.~ Gestion_moment_5, data=Estim, FUN= "sum")
Estim$S_Pl <- specnumber(Estim[2:50])
Estim$Ab_Plant <- rowSums(Estim[2:50])
Estim$S_Poll <- specnumber(Estim[51:149])
Estim$Ab_Poll <- rowSums(Estim[51:149])
Estim <- Estim[, c(1,150:153)]
Estim$Pl_tot <- Estim$S_Pl/49
Estim$Poll_tot <- Estim$S_Poll/99# Plantes : 49 sp au total dans les quadrats
# Pollinisateurs: 99 sp au total dans les quadrats
shape <- c("Plantes" = 8, "Pollinisateurs" = 19)
Estim %>%
ggplot(aes(x = Gestion_moment_5)) +
geom_point(aes(y = Pl_tot, color = Gestion_moment_5, shape = "Plantes"),
size = 3) +
geom_point(aes(y = Poll_tot, color = Gestion_moment_5, shape = "Pollinisateurs"), size = 3) +
scale_shape_manual(values = shape) +
scale_color_manual(guide = 'none',values = c("Graminées" = "#fbcb09",
"Fleuri" = "#ff7207",
"Semé" = "#de1e21",
"Tonte récente" = "#6abe1d",
"Tonte tardive" = "#2b790c")) +
labs(x = "Type de gestion", y = "Proportion d'espèces relative à la diversité totale", shape = "") +
# pour chaque type de gestion sur le nombre total\n d'espèces rencontrées sur l'ensemble des quadrats
theme(legend.position = c(0.9,0.95),
legend.text = element_text(size=10),
axis.text=element_text(size=10),
axis.title=element_text(size=14))